For any given thing or category of thing, a tiny minority of the human population will be enthusiasts of that thing, but those enthusiasts will have an outsize effect in determining everyone else's taste for that thing. For example, very few people have any real interest in driving a car at 200 MPH, but Ferraris, Lamborghinis and Porsches are widely understood as desirable cars, because the people who are into cars like those marques.
If you're designing a consumer-oriented web service like Netflix or Spotify or Instagram, you will probably add in some user analytics service, and use the insights from that analysis to inform future development. However, that analysis will aggregate its results over all your users, and won't pick out the enthusiasts, who will shape discourse and public opinion about your service. Consequently, your results will be dominated by people who don't really have an opinion, and just take whatever they're given.
Think about web browsers. The first popular browser was Netscape Navigator; then, Internet Explorer came onto the scene. Mozilla Firefox clawed back a fair chunk of market share, and then Google Chrome came along and ate everyone's lunch. In all of these changes, most of the userbase didn't really care what browser they were using: the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family.
So if you develop your product by following your analytics, you'll inevitably converge on something that just shoves content into the faces of an indiscriminating userbase, because that's what the median user of any given service wants. (This isn't to say that most people are tasteless blobs; I think everyone is a connoisseur of something, it's just that for any given individual, that something probably isn't your product.) But who knows - maybe that really is the most profitable way to run a tech business.
"Shoving content into the faces of an indiscriminating userbase" maximizes eyeball time which maximizes ad dollars. Netflix's financials are a bit more opaque but I think that's the key driver of the carcinisation story here, the thing for which "what the median user wants" is ultimately a proxy.
Likewise, all social media converges on one model. Strava, which started out a weirder platform for serious athletes, is now is just an infinity scroll with DMs [0]
I do however think that this is an important insight:
> This isn't to say that most people are tasteless blobs; I think everyone is a connoisseur of something, it's just that for any given individual, that something probably isn't your product.
A lot of these companies probably were founded by people who wanted to cater to connoisseurs, but something about the financials of SaaS companies makes scaling to the ad-maximizing format a kind of destiny.
> "Shoving content into the faces of an indiscriminating userbase" maximizes eyeball time which maximizes ad dollars
I mean that's not really the case for paid services without ads like Netflix. They lose money the more you watch. Ideally you'd continue to pay for the subscription but never watch anything.
>"Shoving content into the faces of an indiscriminating userbase" maximizes eyeball time which maximizes ad dollars. Netflix's financials are a bit more opaque but I think that's the key driver of the carcinisation story here,
Non sequitur. For the longest time, Netflix had no advertisements. Do they even now? (I don't subscribe... all their shows end up on my Plex anyway.)
Most great products start out for enthusiasts and often by enthusiasts. They’re opinionated, sharp, sometimes rough, but exciting.
Then VC funding comes in, and the product has to appeal to a broader audience. Things get smoothed out and the metrics rule decisions.
Eventually, the original enthusiasts feel left out. The product’s no longer for them.
So a new product comes out, started again by enthusiasts for enthusiasts. And the cycle repeats - unless someone chooses to grow slowly and sustainably, without raising, and stays focused on the niche.
You start by innovating a "fast horse". This gains you early adopters who pull in a larger audience. A horse can only be so fast, so continued innovation might lead to something more like a car. This will only cause you to bleed users. Stick to the horse.
Instead of continuing to innovate endlessly, you switch to exploitation. Fire the visionaries. They're just a waste of payroll. Bring in people who can squeeze every last dime out of your user base.
-----------------------
The above isn't anything new. However, it's clear that some companies are better at maintaining quality while exploiting. Are they doing something different, or is it just that their customers have to choose them repeatedly? e.g. Most people don't sign up with one car company for life. They'll buy several cars over their life and that's a choice that the car company must win each time. Meanwhile, people sign up for Netflix or Spotify and stay subbed. They don't look at the alternatives every few years. Porsche needs to keep up with the latest and fastest horses to continue exploiting their reputation, while Netflix can focus purely on making more money from their users. A faster horse may come along, but Netflix doesn't break down and need to be replaced.
Can can git rich by growing slowly in many cases - but it will be a long hard road. You could instead sell out today and get rich instantly.
If you start the slow growth path at 30 and retire at 65 you will overall make more money from that thing vs someone who sells out at 35. There are some catches though. The person who sells out can go on to the next thing which in sum total may be more sell out enough to make far more over their lifetime, while the slow growth plan you are stuck. The slow growth is over very slow at first, you often spend 10 years making far less than someone who is "working for the man", then 15 more years more or less even, and only then start making good money. There is no guarantee that you will be successful, some people spend their entire life making less than they could "working for the man"; others go bankrupt when a new VC competitor suddenly gets better by enough to take your customers.
There is no right answer. VC money sometimes is the best answer - but many people who reaching for VC money when their better long term answer would be to grow slow.
Thank you, that makes so much sense to me. Spotify has, for quite a long time, seemed like a product for people who want to hear music, but don't really like music.
It hadn't occurred to me that that might actually be exactly what's happening.
Some very important things get better because of the mass market and investor dollars. iPhone/Macbook are the canonical example.
The hard bit is to keep taste and discipline at the forefront of design. To not let short-term thinking pollute long-term ambitions. Easier said than done.
right, all that and increasing regulation and enforcement, ( SAFTEY SaFTEY SaFETy, agggghhhhh)
marginalises, and criminalises anyone looking for
something out on the edge
and the edge gets crazyer.....think , the street raceing/drifting sceen, where, somebody gona die
and nobody much cares, it's way too fucked up too even make a movie on
the real mofo's and mofo'ets, are working as "contractors", anything goes, again...no movie's or branding possible
at the other end are hard core solder iron in hand
hackers, ocd'ing on PWNE'ing everything in sight
And with my own fucking eyes, I have seen amish boys in town, whipping there horses into a frenzy as they drag race there buggys down main street, not making a movie on that either, cant brand it, it's all thats left.
The market is starving for something authentic, but, every single thing is stolen, branded, comodified, and wrung dry as fast as you can spit
so we get small legions of people who have fetishised things like listening to white noise
> Then VC funding comes in, and the product has to appeal to a broader audience. Things get smoothed out and the metrics rule decisions.
> Eventually, the original enthusiasts feel left out. The product’s no longer for them.
I am immediately reminded of when Slack got rid of markdown-style inline formatting, in favor of a WYSIWYG interface, and the internet (or at least, the corner I live in) collectively (and, imo, correctly) lost its shit at them.
I think this is just taking the common-on-HN refrain of "boostrap good, VC bad" and trying to apply it to everything. I don't particularly agree with the refrain in the first place, and definitely don't think it applies so broadly.
There are many reasons for products getting worse; sometimes it's because the company making it is trying to appeal to a broader base, so the original enthusiasts are no longer the target. Sometimes it's because different people are involved and they can't produce to the same level. Sometimes things don't really deteriorate, but new innovations make other things more appealing.
Since more of our culture is in online, advertising-dominated spaces -- the forces of capital have a lot of incentive to ensure smooth growth straight to the sociopath phase.
Maybe the key is just accepting the cycle of it all and ensuring there's always cool new places for creative/enthusiastic people to do their thing.
This is far too generous a story. While true in certain basic respects, it is a process of capitalists changing products to favor their own interests over users in a mature market. It is a process that begins with the promise of individual empowerment (in a new and growing market where companies are forced to appeal to customers) that ends in a kind of silken chains (when the market is well understood and companies are optimizing their financials).
Notably, this process usually involves not creating a simplified interface that can be turned into expert mode but actively removing features, adding cues, and steering user behavior though a psychological maze to achieve desired effects.
Basically, people should understand this corporate lifecycle and stop being deluded by the opening moves in a new market that superficially appear to favor customers, individual empowerment, etc. It is a process that always ends in heartbreak because it serves investors, not the public.
What you are describing is explained beautifully in "The Tyranny of the Marginal User" essay that got a lot of commentary on HN previously, https://news.ycombinator.com/item?id=37509507.
My favorite quote ("Marl" is the hypothetical name for the marginal user):
> Marl’s tolerance for user interface complexity is zero. As far as you can tell he only has one working thumb, and the only thing that thumb can do is flick upwards in a repetitive, zombielike scrolling motion.
For street usage, I think those cars are popular because they’re beautiful more than because they’re fast (or because enthusiasts like them).
My utterly soulless Lexus will drive more than fast enough to get me in serious trouble. No one will look at it and feel stirred by its beauty, whereas the typical Ferrari or Porsche coupe will look at least appealing to most and beautiful to many, even those who can’t tell the three marques apart or even unaided recall the name Lamborghini.
But a large portion of their beauty is reflective. The Countach was seen as a very ugly car by many when it was released. But it was lust-worthy for its performance. That lust-worthiness over time transformed the car's image, and now it's seen as iconic.
I agree, and I feel the beauty oftentimes comes from the intrinsic love evident in the machine. Looking at a Ferrari it's evident Enzo had a passion for autos. This can also cross boundaries (eating at fine dining restaurants, fine art gallery layouts, etc.) and is probably discernible in MOST things people put out.
That doesn’t explain why japanese manufacturers who used to make sports cars in the 90s don’t anymore.
It’s a mixture of enthusiasm and conspicuous consumption. Most enthusiasts love 90s japanese cars, but the average person sees an old mazda and recoils.
But put an old ferrari in front of anyone and they have a completely different reaction.
Honest question: are you not a "car guy/girl"? Lexus people absolutely love Lexuses. I recently sold mine (needed something larger after having another kid) and I miss it every day.
It always surprises me how quickly people forget nuance.
OK so we're discussing niche vs mainstream, or "what most people want" vs "what a few want".
The few cars you listed are not popular in the ownership sense, but they are well-known and aspirational.
People can buy them to show off status / money / exclusivity, or perhaps beauty. Speed is table stakes, of course. They have to objectively be better than most cars but also special. They can be strikingly beautiful or strikingly hideous but they must not be ordinary.
If you watch / read reviews of those cars, then it tends to be from the enthusiast driver point of view. Is it good at racing, cornering, reading the driver's intentions and reacting instantly and accurately? But then more often than not, those that can afford them do not buy them to use them for that purpose (or at least not frequently.) Many are treated a bit like investments or merely items in a collection.
What a long-winded way to get back to the original point of faster horses and enshittification of software, eh?
Netflix and Spotify might as well be a Toyota Corolla or Prius. I lost my train of thought. I think I just wanted to pontificate about exotic cars for a while.
(I drive a Polestar 2. It looks like a Volvo, is heavy as a dump truck, but damn is it fast as hell.)
That is exactly what is happening to Reddit. Made famous by its submitters and moderators. Business decision driven by metrics based on view counts because that sells ads. Let this be a lesson: metrics are not the only way to measure success. I worked at a company where metrics were viewed as a way to cut through dissonance and bias. Newflash: leaders should be opinionated and have visions that do not yet exist. They should be investors in their product and its culture. Metrics should play a role in that decision, but perhaps a tiny one. Because what metrics you choose, how you measure it, and most importantly, what is even measurable, have a tremendous impact on the effect of those metrics.
You keep using the word "should", but what makes you think these business parasites aren't getting exactly what they want by making their products complete garbage? The CEO caste doesn't care about making good or unique products; they don't care about their users; they don't care about company culture; they don't care about their effects on society or the environment; they don't even care about the long-term financial success of their company. They only care about the immediate short-term gains that directly benefit them, and clearly paint-by-metric is a tried and true way of optimizing for that at the expense of everything else. If it rots the company from the inside out (or even society as a whole), who gives a shit? They just fly off and find a different company to parasitize.
By the time our society is collapsing and our rivers are catching fire and our government is being overthrown and our oceans are boiling and our bodies are full of plastic and we can't even escape to another planet because of Kessler syndrome -- all due to their actions -- they'll be old. That will be their kids' problems, and we know the CEO caste fucking hates their own kids.
I get your point but I think the browser analogy is wrong.
IE had something like 90% market share back in the day because it was bundled with the OS and cost $0.
Chrome ate everyone's lunch because everyone was using google to search for stuff, and they could advertise their browser on their home page or together with their search results. They also took out ads, in some countries, on billboards, in newspapers and even in cinemas.
I'm sure technical people talking to their families had a small effect (though wouldn't they recommend firefox, because FOSS?), but I think that pales in comparison to google being able to advertise chrome on their search page.
>Chrome ate everyone's lunch because everyone was using google to search for stuff, and they could advertise their browser on their home page or together with their search results.
That and it was such a better browsing experience. Firefox was not good compared to Chrome for years. I'm sure they are feature parity now, but for years the Chrome experience was significantly better.
You’re giving it way too much of a positive spend. None of the companies are using analytics to increase the desirability for the majority of users.
They are doing it to increase “engagement” and so more people will stay on their site longer.
Why else wouldn’t Netflix show the “continue watching” row first instead of forcing you to scroll past algorithmic generated crap?
It is the same reason that Google went from describing success as people getting off their site faster and going to one of the “ten blue links” to the shit show it is today.
which is also what I feel about the Spotify algorthim at times — no matter what I'm listening to, it invariably brings me back to what it thinks are my "old reliables" once it gets onto recommending stuff.
I might just listen to it, if I have it on in the background, which then in turn feeds the algorithm that it made the "correct choice", but it's a million miles away from, say, listening to a radio DJ where you like their rough output but they're cherry-picking what to play next.
To this point, I've been using Qobuz as an alternative and it's recommendation engine is laughably bad, but the experience is somehow better. I'll get the most random songs pop up in the list, and sometimes it's a very pleasant surprise.
In the world of music discovery a bad recommendation engine is maybe better than a hyper-fine-tuned one.
> if I have it on in the background, which then in turn feeds the algorithm that it made the "correct choice"
I have a very horrible case of this. One day at night, I slept listening to lofi playlist. The next week all my recommendations were screwed. Horrible assumption on the part of algorithm.
I’m experiencing this in Peloton-land. They have an app that purports to be for home gym enthusiasts but is actually optimized for people who want to take instructor-led classes on their phone. Certain features don’t work as advertised and I quickly reasoned that while this is a pain in my side most users don’t care. If they did, Peloton would fix it.
You’re way overestimating the effect an enthusiast has. Evangelism only goes far enough to introduce people to the thing. How often someone uses the thing depends entirely on its utility (usefulness).
As long as Netflix was successfully reading the author’s mind, they were satisfied with the experience. However, Netflix assumed that they want to keep watching the same content, oblivious to the author’s desire to discover something entirely new. Netflix failed to meet the expectations of those seeking something entirely different.
I can understand why Netflix made this change. They’ve replaced many shows with their own in-house productions. By doing so, they prevent users from searching for specific shows and then realizing that Netflix doesn’t have them. If this happens frequently, they risk losing customers.
On the other hand, Spotify doesn’t face this issue. Therefore, I’m puzzled by why they’ve made it more challenging to explore content by categories. (Disclaimer: I don’t use Spotify, so my experience is based solely on author’s observations.)
Luxury watches are a good analogy too. A $5 watch from the gas station will give you the time just fine but there’s a market for watches costing hundreds of thousands of dollars.
You don't even need a watch. Smartphones can tell you time (you can configure to show times for many timezones) yet there is a market for watches (luxury or normal)
> This isn't to say that most people are tasteless blobs; I think everyone is a connoisseur of something, it's just that for any given individual, that something probably isn't your product.
I think this is a great nuance that is often overlooked when discussing this.
Nice example, but not everything is like automobiles where probably not even one in 1000 people has ever been to a track day let alone actually raced a car, but sporty marques are desired.
A very large portion of people actually cares about what they are searching for, and want the ability to ACTUALLY search and find that, with real parameters, not merely get some not-even-close stuff shoved onto their screen instead. That is NOT the serendipity of browsing the stacks in a great library.
A great example of failure is Amazon. I run a small design & manufacturing business, and years ago started getting pestered by Amazon about "Amazon Business" trying to supply both office staples and parts to businesses. This was an area that had enormous potential. Yet, they have entirely failed. I've never bought a single item, and it has faded.
Their primary competitor is McMaster-Carr [0] who does it right. Well-defined categories of everything, and highly specific search capabilities, at reasonable but not bargain prices. EVERYTHING you might search for is fully parameterized in every dimension and feature. Min/max/exact, width/depth/height/thread/diameter/material/containerType/etc./etc./etc. appropriate for each type of product. The key is McMaster DOES NOT WASTE MY TIME. I can go there, quickly find what I want or determine that they don't have it, and get on with my day.
The smaller company that does it right is still beating the tech giant a decade later. Same for other similar suppliers who actually have a clue about what their customers really want.
They continue to prevail over tech giants and VC-funded sites BECAUSE THEY ARE NOT STUPID.
It would be nice if the tech/vc crowd would also stop being stupid. They started out not stupid, but they really lose the plot when they think a few extra eyeballs this week will really win in the long run. At least provide two modes, a strict and serious search and their new messy UI. But they are stupid and this will not happen. Enshittification rules the day.
The thing that really pissed me off about Amazon Business is that they bought Small Parts and killed it off. Small Parts was a tiny version of McMaster-Carr that specialized in fasteners, small diameter fluid handling, short sections of specialty materials, and in general, quality "small parts."
If I bought directly from Small Parts, I knew I'd get exactly what I wanted. Ordering from Amazon Business? A complete crapshoot.
Going to www.smallparts.com now just redirects to an Amazon 404 page!
I've long wondered why Amazon made it harder to buy products from them, why they've decreased the [customer] value of their search, decreased the value of the filters, decreased the value of the reviews...
I mean the answer has to be "they make more money this way" but for me it's means I groan internally before going to Amazon because finding the product I want will be almost impossible - it's even hard if I already visited and already found what I wanted to buy, finding it again, near impossible. Not even basics like search by product manufacturer actually work.
> is the most profitable way to run a tech business.
Yes, I agree. This does seem to be the most profitable model for running a tech business: maximizing user engagement or increasing the time users spend on the platform. Whether that’s achieved through intentionally convoluted UI or by aggressively surfacing certain content, the end goal remains the same.
That said, I don’t think there’s much room left for significant innovation in video streaming interfaces. The core challenge continues to be content — whoever offers the best or most compelling library wins. UI changes might tweak engagement metrics by a few percentage points, but they’re marginal compared to the impact of strong content.
At the end of the day, if there’s a great movie or series to watch, people will show up. If the content isn’t there, no amount of clever interface design will convince someone to spend 30 minutes on something they’re not actually interested in.
> However, that analysis will aggregate its results over all your users, and won't pick out the enthusiasts, who will shape discourse and public opinion about your service. Consequently, your results will be dominated by people who don't really have an opinion, and just take whatever they're given.
> In all of these changes, most of the userbase didn't really care what browser they were using: the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family.
I'm confused as to whether your saying change is caused by catering to the median who doesn't care, or the enthusiast who recommends the latest and greatest. You seem to be saying both.
Yeah, I could have been clearer there. The browser developers started by catering to the enthusiasts, who switched. The enthusiasts then told the majority of the userbase that the new thing was better, and so the majority switched, causing the large-scale changes in market share.
This is exactly the situation unfolding with JetBrains right now. They've lost all touch with their professional enthusiast core and have gone hell-bent on acquiring new users at the cost of alienating a big chunk of their core. I don't think it's going to go well for them, they don't have the chops to compete with Microsoft like they're presently trying to do.
I think the second paragraph in the parent comment fits really well with mimetic theory and this René Girard quote: "Man is the creature who does not know what to desire, and he turns to others in order to make up his mind. We desire what others desire because we imitate their desires." This, however, doesn't mean that the current Netflix solution is the only one possible.
> if you develop your product by following your analytics, you'll inevitably converge on something that just shoves content into the faces of an indiscriminating userbase, because that's what the median user of any given service wants
Except you're making the mistake of thinking these services are optimizing for their userbase.
They are not.
They are optimizing for revenue and profit growth,
a very different target.
More ads,
cheaper and easier-to-product content,
lower opex.
They are converging to churning out the least offensive slop at the cheapest cost with the maximum revenue.
None of the analytics are about what people using the product want,
they are about making the most money and growing the fastest.
Nothing would look like the services mentioned in the article if they listened to what the users really preferred.
this is such a fantastic comment because it makes a charitable attempt to explain how data driven decisions go off the rails.
and it matters because this seems to be an omnipresent phenomenon.
everything everywhere seems driven by this unless someone with decision making power is executing a specific and conscious strategy that pushes back against it.
>maybe that really is the most profitable way to run a tech business.
That's the issue, it seems like it really is the most profitable way to do things. Everything sucks now because shooting brainrot and advertisements at our eyes and ears is more profitable than actually giving us what we want.
Just for protocol: In your example the peasant rabble recommend Chrome, while the enthusiasts with deep technical knowledge, broad perspective, wisdom, charm and good looks recommended Firefox.
Then, how could a business identify its (or market's) trend-setters, enthusiasts, or whatever we call them, which will push towards something new? I see this as essential for either making the business better, shinier, or to avoid losing users.
By participating in the community. Content moderation on HN is so much better than on Facebook because dang is one of us, whereas on Facebook, it's a team of people in a developing country, in a different cultural context. Netflix needs to be run by film enthusiasts, not UX engineers trying to disguise the fact that all the good IP has been pulled back to the streaming platforms of the original producers. Spotify needs to be run by music enthusiasts, not people pushing covers of pop songs to avoid paying royalties to the original artists. And so on.
Indie Hackers is full of people trying to flog their shit AI-powered marketing SaaS, because they've never done anything other than software engineering, so they don't know any good problems to solve. There are uncountably many good problems out there, each with thousands of people who would pay you money to solve them, but those people don't know their problems can be solved by a computer, so you have to go out into the world to find them yourself.
Just like football scouts need to actually visit some niche teams and watch not that interesting stuff to find talent before it is too late.
With tech it might be easier because you might create niche groups so those people come to you.
Just like PG created HN. Nowadays HN is too mainstream so all ideas here are seem already popular so it is like going to scout high school t am that won local championship everyone already knows which players are lined for pro contracts.
By risk taking on good ideas rather than always trying to pivot your way from the status quo.
Product-Market fit is great if you're developing a SaaS business but it's not necessarily going to give you new inventions — something new is speaking to a potential gap in the market that doesn't currently exist.
Teams should identify their drivers of key metrics and do power user analysis based on this. A halfway decent analytics team should be thinking this way.
Ultimately, analytics are just a view into the business. This thread is complaining about doctors not using microscopes when diagnosing system issues - sometimes a narrow slice is important, sometimes you need to zoom out. If you focus on your "early adopters" or power users exclusively, without understanding how they affect the business, then you are at risk of building things that most of your user base doesn't want.
yeah except a lot of those companies almost went bankrupt trying to make those cars for enthusiasts and only for them.
porsche and lambo didn't see the outsized success they have now (financially) until they started pumping out SUVs. hell, the purosangue was made precisely to capitalize on that boring market segment.
i feel there's a little suvivorship bias at play here. i think the important thing is to not forget your enthusiasts perhaps, but a lot of these "successes" wouldn't even be around were it not for appealing to the greater masses. ofc some market segments fare better and you can build a business around enthusiasts.
>For any given thing or category of thing, a tiny minority of the human population will be enthusiasts of that thing, but those enthusiasts will have an outsize effect in determining everyone else's taste for that thing.
I think that's a self-dellusion many tech enthusiasts have, that they're somehow trend-setters.
And then the same enthusiasts say for the original iPod "No wireless. Less space than a Nomad. Lame", and see the masses jump to buy it, and themselves only catch up later.
Or they see the masses never caring for their e.g. desktop Linux, whose mass dominance (not mere "works for me" or "have set it up for my elderly parents and they don't even know it's not Windows") would come "any day now" for the last 30 years...
Trend-setters exist, but they're a different group than the "tiny minority" of enthusiasts. More like some musician paid to spot Beats headphones, or some cool dude sporting some gadget early on.
>For example, very few people have any real interest in driving a car at 200 MPH, but Ferraris, Lamborghinis and Porsches are widely understood as desirable cars, because the people who are into cars like those marques.
A hell of a lot of people had a real interest in driving a car at 200 MPH, if they could have the chance. And even more admired Ferraris, Lamborghinis and Porsches because of their design and elegance (and price, people aspire to luxury goods, even when they can't afford them), not because some sport-car afficionados said so.
It's the same in other areas: the popular books, or comics, or movies, or music, etc. are rarely if ever what the "inner" crowd of each niche admires. Most people buy Reacher and such, not Finnegan's Wake.
>So if you develop your product by following your analytics, you'll inevitably converge on something that just shoves content into the faces of an indiscriminating userbase, because that's what the median user of any given service wants.
More likely, if you want to keep and continue increasing your margins, and your stock price, you'll incrementally continue to shit all over your product trying to squeeze ever more money.
Neither the "enthusiasts"/tech-savvy users NOR the "median user" wants Netflix to be the shit it has become, or Google search to be so fucked up, or ads and nags on Windows UI, and so on.
They're just given those, and they accept them having no recourse. The moment there's a better recourse, they jump to it (like IE -> Firefox -> Chrome, or BS early search engines -> Altavista -> Google).
> very few people have any real interest in driving a car at 200 MPH
I agree with that.
> but Ferraris, Lamborghinis and Porsches are widely understood as desirable cars
I agree with that too.
> because the people who are into cars like those marques.
I think that is not true. I don’t care about cars. Never had one. Don’t even have a driving licence.
The reason why i think Ferraris, Lamborghinis and Porsches are desirable cars is because they look cool, and they sound cool. They were designed to be like that. If i see one on the street i notice it. I couldn’t care less about the opinion of gearheads. If a car would come out looking like my grandpa’s skoda, but all the car lovers would love it I wouldn’t even hear about it.
It is all about flashyness of the industrial design. And rarity of course.
By that logic if you start making rare cool looking sportcars they would be automatically desirable. I doubt that would happen. Unless gearheads give it "approval", no one would buy it and you will be out of business.
I think when you're a startup, you have to invest in all of these things - you want to hire some experts early on because they'll have insights that help you design a better product, and if your product appeals to experts it will be a PR win. But of course your goal is scale and distribution so you have to respect a certain lowest common denominator as well lest you become too niche.
Once you become a bloated monopolist like the three companies you just mentioned, your distribution strategy is solved in other ways (like, you've done some bundling and some acquisitions, maybe pressured a few companies into exclusivity agreements and are probably breaking some anti-trust law or other but you have lawyers). Then you don't care about the experts, PR or niches anymore, and you serve up slop. When the analytics recommend slop you go with the analytics, when they don't you ignore them.
None of this is to discount your insightful comment, just saying once you're big enough, your strategy is just doing tricky distribution deals, really (a fact no record executive would dispute).
It's probably profitable in a lot of cases to follow those metrics, shovelware content is cheaper to produce, and since the median user pays the same subscription fee as the enthusiast, you get better margins producing slop for the uncaring masses.
You need enthusiast businesses owners to produce quality product.
Damn, I never thought of this before, but it explains so much!
You’re talking both about tastemakers and the silent majority vs loud minority.
I promise it is NOT always a good idea to follow the enthusiasts, because they are not at all like everyone else who uses your thing. Following them will skew your decisions—unless they are your entire customer base, so, have at it.
This article imo is complaining about the effect of middle management product owners at large companies. There are two dynamics that both converge on enshittification:
1. These product managers (or product designers) are early in their careers and want to make a splash. They are given lower priority projects but try to break out by making them bigger, better, more non-horse-like. They over-design and over-complicate the solutions as a result, because they don’t yet know when the right solution is just a refinement of what’s tried and true. They are incentivized to do this because they want to break out of the mold.
2. The managers above them, or a layer or two above depending on company size, are risk AVERSE. They are tasked with delivering results regularly and consistently. If you have the innovation bug or are creative at this layer, you get moved onto projects where this is required, which is not most of them. Overcomplicated is fine sometimes with you but WEIRD is absolutely not okay (the stuff that actually could be innovative), and no one gets fired for following The Metrics.
These two incentives clash to create overcomplicated but functionally poor products that aren’t helping anybody out. A healthy skepticism of complication and a healthy skepticism of engagement as the sole metric (or metrics in general) is necessary to make good shit. Sometimes it is actually understanding and using things as an enthusiast would, but you need to bring in an understanding of how the rest of your users are distinctly different from the enthusiasts, too. Using your thing yourself and actually following your own subtler feelings is what produces really useful innovation, slowly and surely over time.
Some people have claimed that pure A/B testing is an agent for enshittification, both on a quality and ethical dimension. And I can’t see how those people are particularly wrong.
There are systems out there that can do AB/CD testing and those do a better job of finding pairs of changed that have compounding effects.
You cannot A/B test your way from chocolate and peanut butter to cherry and vanilla. So we get to deal with tone deaf companies who feel their analytics are proving that customers either don’t know what they want or are lying about what they want. But that’s not something A/B testing can prove. It takes more sophisticated experiments than that.
> Ferraris, Lamborghinis and Porsches are widely understood as desirable cars
... primarily for their price tag. There are a lot of enthusiasts for money in the world, much more than for driving at 200 mph.
> the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family
It was never about recommendations. MSIE and Chrome were (and are, but with Edge Browser instead of MSIE) shoved into consumers' throats by ads, marketing, bundled distribution and outrageous lies.
The short term data driven optimizations somehow erode the original product architecture and some of its value. I also think treating the consumer as static. Trick me one shame on you, trick me twice (admittedly I get tricked even more often to click on stuff) shame on me but eventually I learn and what worked turn into a constant irritating torn-off. These irritations accumulate. Good product management should strive to minimize such irritations but I guess we lost that with Jobs.
Something is popular, folks are envious of it, they end up building something much like it. Doesn’t matter if it’s houses, logos, or user experiences – seems to be how things work.
eh, I feel like this is a nicely typed out comment but it hits some wrong notes.
1. I wouldn’t say the car veands you mentioned are popular because they can hit high speeds. In my experience nearly any car can with the right engine and equipment in it (of course due to weight distribution and other details I assume they’re not all equally safe but that aside).
Personally when I look at those brands I think they’re sleek and pretty and when I feel like wanting one it’s because they’re expensive cars, driven by the rich. They’re not chosen only by the rich cause they have the best taste, they’re chosen by the rich because they are the only ones to have the financial means to afford one.
Also I feel like the changes made based on analytics arent made to please (more) users but to make as much money as possible, whether that be pleasing users in the starting phases of your company or in the latter phases when you already dominate the market squeezing money out of your big existing userbase.
> you will probably add in some user analytics service, and use the insights from that analysis to inform future development. However, that analysis will aggregate its results over all your users, and won't pick out the enthusiasts, who will shape discourse and public opinion about your service. Consequently, your results will be dominated by people who don't really have an opinion, and just take whatever they're given.
This is so spot on. I was a long-time serial entrepreneur who spent a couple decades across three successful startups discovering, shipping and growing new categories of tech products primarily for consumer, prosumer and hobbyists. Then I sold my last startup to a very large F500 silicon valley tech leader and ended up a senior product exec there. While there were a lot of positives like more mature engineering processes, testing and devops as a discipline, the exact issue you describe was a nightmare of product-damaging mistakes I called "analytics abuse." In my startups I valued having increasingly robust analytics over the years. In part because they helped increase my overall understanding of usage but mostly because they provoked good questions to explore. That exploration happened naturally because as the "product guy / founder" I never stopped spending a lot of time with our most passionate, opinionated, thought-leading customers. Over years of iteration I'd learned how to engage deeply and listen carefully to input from these customers. This involved interpreting, filtering and curating the mess of divergent personal preferences and pet feature ideas to tease out the more actionable product signals that could increase broad usage, adoption and passion around our products. I'd then bring those curated signals back to the product teams for evaluation and prioritization.
At BigCo they were diligent about meeting with customers, in fact they had entire processes around it, but their rigorous structures and meeting agendas often got in the way of just directly engaging and actively listening. Worse, the customer meetings the more senior product decision makers actually attended in person were mostly with the highest revenue customers. Junior PMs (and sometimes new grads) were delegated to meeting with the broader base of customers and filing reports. Those reports were then aggregated by ever-helpful program managers into tables of data and, eventually, slides - losing all nuance and any ability to spot an emerging outlier signal and tug on that thread to see where it goes.
I tried to convince everyone that we were missing important customer signals, especially from our smartest, most committed users. Being only one level removed from the CEO and quite credible based on prior success, I was definitely heard and most people agreed there was something being lost but no one could suggest a way to modify what we were doing that could scale across dozens of major products and hundreds of product managers, designers, execs and other stakeholders. In my experience, this general problem is why large companies, even the most well-run, successful ones full of smart people trying their best, end up gradually nerfing the deeper appeal in their own products. Frustratingly, almost every small, single step in that long slide pushes some short-term metric upward but the cumulative effect is the product loses another tiny piece of the soul that made our most evangelistic, thought-leading customers love the product and promote it widely. Ultimately, I ended up constantly arguing we should forego the uplift from some small, easy-to-prove, metric-chasing change to preserve some cumulative whole most people in the org weren't fully convinced even existed. It was exhausting. And there's no fighting the tide of people incentivized on narrow KPIs come bonus season.
I'm sorry to report I never found a solution to this problem, despite my best efforts over several years. I think it's just fundamental. Eventually I just told friends, "It's a genetic problem that's, sadly, endemic to the breed" (the 'breed' being well-run, very large tech companies with the smartest product people HR can hire at sufficient scale). Even if I was anointed CEO, given the size of the product matrix, I could only have personally driven a handful of products. I do think codifying premises and principles from the CEO level can help but it still gets diluted as the number of products, people and processes scales.
Given several mrandish-equivalents, gathered into a side-channel Customer Advocacy org, is there some way to integrate their output without this problematic constantly arguing against metric-chasing?
I'm groping towards something vaguely ombudsman-y, or WW2 production/logistics trouble shooters. Or maybe even pre-Bush41 ARPA Project Managers - term-limited person-with-a-checkbook and few accountability constraints.
If one accepts this role has to be out-of-band, vs poking big hairy blob in hope of creating and maintaining signal channels with particular properties, and grants CEO-adjacent leverage, then it seems a remaining unresolved challenge is integrating the output signals at scale? If so, maybe (jest) CA granted KPI offsets?
The irony is that he argued for a faster horse and that’s what all his providers are doing. TikTok is the faster horse. What he really is asking for is a step out of the paradigm, although he argues for a romantic conservative product instead of an innovative product like Ford.
There's a fundamental reality that shapes both Netflix and Spotify's trajectory: content licensing. 2012 Netflix had access to vastly more of everyone else's library, so it was closer to an indexed search of what was available that one could watch and then getting that video onto your screen. Over time, other companies understood that they were underpricing their content and Netflix was reaping the benefits. Once external forces adjusted, the TV/film bidding wars began. Today, netflix doesn't have nearly as much content as they used to have.
That risk (losing all content and facing extinction) is what pushed Netflix in the direction of being a content-producer, rather than a content aggregator. I agree with everyone's points on the influence of the median user in diluting the quality of the content Netflix produces, but that's not the only forced that pushed us here. Spotify faced a similar crossroads and decided to broaden beyond music once they started losing bidding wars for licensing.
Being a faster horse wasn't an option available to either Netflix or Spotify; there is no path for a 'better 2012 version of netflix or spotify' in 2025. They each had to change species or die, and they chose to keep living.
Apple Music still offers library management, with their entire catalog to choose from. They try to play all sides, with algorithmic playback, radio, add to library, and playlists. Adding to library and playlists do seem to be core features, but I’m curious how many people put in the effort when it’s not explicitly required.
On the other hand they are sometimes bad keeping content matched when you add an album to your library and, I assume, the distributor replaces the album with a different version. This also happens with "matched content" when you added a ripped version of music you own.
It’s less obvious than with Netflix because the songs don’t completely disappear. Spotify pays different rates for different songs depending on the label so there are certain songs they’d rather you not listen to and other content it’s much cheaper for them if you listen to so they push you to that content.
> They each had to change species or die, and they chose to keep living.
Did they, though? 2025 Netflix is extremely close to having a worse UX than piracy, and it's already far more expensive. Are people going to pay a fortune for Netflix when their handy nephew can hook them up to his far superior Jellyfin instance for a sixpack of beer?
It's a tragedy of the commons, really. The whole value is in having a complete catalogue available for the casual viewer, and making $10-$20 from someone wanting to watch a random decade-old movie twice a month or so. Break up that catalogue into twenty different services each charging $15, and that same casual viewer isn't going to subscribe to a single one of them.
If the streaming industry doesn't get its shit together they are either going to lose viewers to piracy, or to a completely different medium.
You are overestimating how many people have access to a piracy nephew by a very, very large margin. And even if we all knew a privacy nephew, they're very quickly going to stop responding to incessant requests for more content. And they won't be available 24/7.
The TikTok-ification of advertising supported platforms is terrible, but makes sense to me. LinkedIn pivoted from making money on subscriptions and fees for job postings to ads, which mean the leading drivers are 'engagement' e.g. time you spend doom scrolling on their platform. This will end in disaster for the platform as a place to find jobs or employees.
Netflix I understand much less. They make money from subscriptions. If you perceive having a fantastic experience on the site by just going there, finding something you enjoy watching, and leaving... they win. Why they would foster a doom-scrolling experience I really can't really explain, other than imagining some dark pattern like they have to pay per view and want you to watch C grade movies? More time spent looking for something to watch means less time streaming?
I assume it's about papering over the gaps in their content library.
You can't provide a seamless UX for turning on the TV and watching The Office if you don't own the rights to The Office. They want to habituate you to scrolling through content Netflix actually owns and picking something, because it's apocalyptic for them if you ever treat the services as fungible content libraries that you hop between month-to-month.
A short while ago, I noticed I only used Netflix to watch 2 classic comfort shows, and I started to doubt if it was worth a 2-classic-comfort-shows-as-a-service subscription. I tried looking through the catalog to see what else I was paying for and ended up cancelling my subscription.
Netflix does an amazing job in giving the impression that they have an endless library of top quality content, but in reality, it seems like it's only a handful good shows and some filler, but presented in a way that makes it look like there's way more than it actually is.
This. The absolute worst case scenario for streaming is you open the app scroll for a minute or less then close it. If you scroll for 10 minutes instead of just 1 the streaming service has much larger mindshare and youre more likely to check again tomorrow.
> More time spent looking for something to watch means less time viewing?
or, if you're presented with more random 'clips' or movie snippets, this turns on your gambling reward center. It's like a slot machine - where you "win" by finding a good series to watch after searching. And because this is random, you end up getting addicted to looking thru the list/snippet, trying to encounter a perfect series to watch.
This is strongly in tin-foil hat territory but: streaming video costs a lot more money than streaming some JSON to populate a UI. Every minute you spent browsing the catalogue over playing a video is probably a significant costs saving for Netflix.
> Why they would foster a doom-scrolling experience I really can't really explain
They want to take the bargaining power from creators (and old IP owners).
They don't want the customers to search for a specific show. They want the customers to watch whatever is shown to them. This way Netflix will have tremendous power over show creators - if our algorithm doesn't favor you, it doesn't matter how good your show is or how much money you spend on marketing outside Netflix.
You've got it backwards, Netflix doesn't want people to just doom-scroll, the users want to doom-scroll.
Attention destroying apps reduce the long term focus and reward centers such that doom-scrolling through the catalog probably feels better than just watching something. Most of the folks I know who start a movie or show immediately pull out their phones anyway to scroll elsewhere.
Because my netflix subscription is cancelled specifically because the "Finding something I want to watch drains my energy" phenomenon. Gradually over the course of like a year I got more and more frustrated with being suggested things, and not having a good way to find things.
I wish Netflix and other streaming services had more information about a movie or show for me to base my decision on. I would like more text. Maybe, some reviewer snippets. The full major cast members, not just the top names. The director should be prominently displayed. Let me easily see what else that director has done, even if it's not on that streaming channel.
Apple TV is the worst, because it dumps you right into the program, and you have to back out in order to get more information.
They all just want me to trust them that I'll love it. I end up having to pull up reviews on my phone.
> Why they would foster a doom-scrolling experience I really can't really explain
Because regardless of whether or not the business model depends upon it, investors have been trained that “engagement” is inherently good quality for their investments to have. Increase engagement, stonk price go up.
Maybe it is winning despite what Netflix leaders are choosing to do, and maybe their choices will cause them to falter soon. And maybe Netflix could be doing better than they are. But it is always easier to pontificate than execute.
I don’t buy Netflix solely because they don’t integrate with the search in the iOS/macOS TV app.
Unfortunately, based on media trends before streaming and Netflix was a thing, lots of people like C grade productions. If you recall, “reality” TV shows were taking over in the 2000s. People like the Tiktok-ificiation (or otherwise lowering of quality).
Netflix was changing a lot to drain more money out of users recently, which is why income rose recently. What I'd like to see is active / recurring users instead.
Mostly it's to cover up that the catalogue isn't as great anymore, isn't it? Since almost every big label took back the rights and started their own streaming service, Netflix simply doesn't have as much content (that anyone would want to see) anymore.
I quit all those platforms recently and I'm not missing the frustration of having to 'switch channels' through their incomprehensible categories and views anymore.
Think of it this way, the less time they spend actually WATCHING content, the longer they will pay their monthly service because they have this massive "watch list" that they never actually get through. They just keep paying month after month never getting through a backlog that they inspire to watch. I don't agree with it, but it makes sense to me. If you can never feel satisfied, you will pay over and over again chasing that satisfaction of watching "everything."
Many people will pay Netflix for years hardly watching content for months just because the convenience factor of not having to subscribe/unsubscribe when they know a new season of X will be out in the next year. It's wild to me, but people are lazy. So again, the more you keep them from actually watching the content and realizing they are "done", the longer they likely just keep their subscription active. Get them to add as much potential content they want to watch to a never ending backlog watch list.
I guess my thing with LinkedIn is that there's just no reason to use the feed. It's still a place to connect with people I've worked with and keep up with what they've been doing. It's incredibly useful for that. I really don't find the feed to be either a boon or a hindrance in that use case. I know it's there, I know it annoys some people, but it's just irrelevant to me.
The LinkedIn feed would actually be very useful if it only showed my contact's milestones such as job updates, their own product/service launches and events like conferences or conventions that involve them.
Of course, such a feed would take me 2 minutes per week to read through so that wouldn't be good for the business.
I think Netflix faces the problem that measuring the causality between a user watching specific content and choosing to stay subscribed is super hard. Therefore, they focus on a metric that is easy to measure, namely time spent in the app. This is likely not the metric they should be optimizing for, but since they _can_ measure it, it becomes the target anyway.
Netflix's primary goal used to be to attract new subscribers. Now it's a more about maintaining subscribers and finding new ways to monetize the existing subscriber base. That's why you're seeing things like "sharing" subscriptions, and advertising, and premium plans.
> Why they would foster a doom-scrolling experience I really can't really explain
Entertainment is a zero-sum market. More time spent doom scrolling means less time spent on another service, which probably reduces their churn (also, ads)
i think people's view of netflix's business model is heavliy biased by what they want netflix to be.
i get it, i hate what they've become too. i'd like to believe there's a world where paying for content is a better model than selling ads. but the reality is that every time netflix makes a decision that the internet gets angry about, their balance sheet looks better.
The thing about the situation is, now that when Tik-tok-ification has grown big enough, it (no-choice interfaces, "enshitification", etc) becomes the only paradigm UI designers, managers and investors understand. Moreover, it's interface that essentially completely controls the user - all the choices they have are essentially fake and control always appeals to managers and control may not immediately make money but it can make money long term so it can be justified.
You can see how Sonos enshitified their interface and even with a user rebellion wouldn't back down, just as an example.
I've heard this called the "Tyranny of the Marginal User".
To keep the line going up, platforms have to appeal to wider and wider swaths of a population, eventually lapping at the shores of a population that really doesn't care or want this service. But if you can hook them with some dopamine in a 5-second video, or a quest to rediscover some neat thing that they saw two page-loads ago but is now mysteriously gone from the very same list it appeared in, then you've clawed one additional user into your metrics and the VCs give you a treat.
These people don't care about the service and they're the worst users to cater to, but everyone caters to them because they're the only ones left. Hence, TikTokization.
The implication here is appeal to a wider audience _at the expense_ of the existing customer base, right? Otherwise it wouldn't be a tyranny at all.
What I don't get is at some point the marginal user increase for a change has got to be smaller than the number of customers you tick off and lose by changing things.
Is the idea that all services converge on the same N billion people target audience who wants something almost entirely unlike the initial product? I feel like "marginal" doesn't really capture this nuance if so.
Existing userswon't leave when you tick them off, if you lock them in somehow (say, by having their social network all using your platform, or maybe by having previously sunk lots of time into curating a library of their favourite X on your platform).
I finally know what to call these idiotic trends that I've learned to recognise but couldn't name.
The one that grind my gears the most has been Microsoft breaking decades-old Windows paradigms to cater for Linux-developers-on-Windows, which is a very marginal, even actively disinterested group. All this at the expense of the 99.9% of their loyal user base.
For example, VS Code had the opposite shortcut (literally with the arrow keys going in opposing directions) for "go back in search history" to every other editor ever made for Windows... but matching the Linux equivalent.
Similarly, they recently broke "cls" to match the broken(!) behaviour of "clear" in Linux because of basically just one or two Linux users complaining in a GitHub Issue ticket. Windows users weren't listened to, because they're already users, not potential new users.
oh, that's extra dumb, since there's no canonical[1] Linux desktop. that said they should offer shortcut presets, like JetBrains products. changing the default is just whack.
[1] oh, what the hell!? pun unintended, but now at least the name makes sense
I also dislike the TikTokification of everything, but I also know that all of us on this platform are wrong in the sense that we're not the user being designed for.
Consumer apps at massive scale like TikTok and Netflix don't design for nerds like us, they design for the average person. Actually, they design for the average behavior of the average person.
And most people on this planet are more or less happy with whatever they're presented with because they don't care about technology.
And when you control what's presented to people, not they (and they don't care), you can push them to consume what you want them to consume.
I heard a YC group partner once that he's worked with a ton of delivery apps. Many of them start out as differentiated apps for ordering from the best "hole in the wall" places or the app for authentic foreign cuisines, only to discover that the best growth hack is getting McDonald's on the app, because that'll be your top seller, instantly.
Most people just do the default thing everyone does—and we're probably all like that in one aspect or another of our lives, and that's who many experiences are designed for.
Overwhelmingly, products are designed to maximize total recurring user interaction, aka engagement or attention grabbing. This is the proxy for ad revenue, the most popular business model (even if Netflix is different). Look at Quora, LinkedIn and even SO, which essentially degraded into content farms for these reasons, largely downstream of the Google search funnel.
But engagement maximization looks the same everywhere – it’s communicating with the amygdala of the user, not their consciousness. And in a way, everyone’s amygdala is kind of the same and generic (sugar foods, violence, rage bait, boobs, chock value etc). Products that are largely designed for higher consciousness are more varied, such as most books. But those drive less engagement.
The amygdala wants more of the same, and the prefrontal cortex seems to want variation. My view is that you can’t have the chocolate muffins and raw carrots on the same plate, or a bookshelf with both Dostoevsky and Playboy magazines. You have to compartmentalize to protect yourself from your own amygdala. Same goes for media. Even well meaning product managers will be completely fooled if they simply follow the metrics.
Yep, totally. Also, much of Netflix's growth now comes from their ad-supported tier, so they're definitely part of that attention economy.
And part of the problem is that if somebody (TikTok) has the most engaging format possible (vertical short-form video) and you (Substack, Reddit, LinkedIn, etc.) don't, you're at a strict disadvantage. So you enable short-form video, boost it in the algorithm, etc. no matter if it's a fit with your product because people will watch it if it's put in front of them.
> My view is that you can’t have the chocolate muffins and raw carrots on the same plate, or a bookshelf with both Dostoevsky and Playboy magazines.
And the problem is that in media, the prefrontal cortex stuff will never make as much money as the amygdala stuff, so few platforms will survive by focusing on the prefrontal cortex stuff.
A big reason HN is still so cozy and surfaces cool articles and discussions is because YC doesn't have to monetize it or optimize for engagement.
One can always do as in "The good place" show: put a bunch of hotties to talk about and play with moral philosophy. I think the show was somewhat evil in that approach, but at the same time, it was also morally sound...
There’s a lot of money to be made in letting people order takeout from McDonalds while not feeling like the kind of person who orders takeout from McDonald’s.
I hate that more than McDonald's. The restaurants making mediocre-tasting but instagramable food in a place that spent more on the interior design consultant than the chef who created the menu.
> Actually, they design for the average behavior of the average person.
They're generally designed for engagement. Nobody is particularly asking for this type of experience it's just that Tiktok has discovered the most addictive - eh hum, I mean engaging - experience thus far. So they're being copied.
Netflix is a little different though as if people open the app and always see the same top titles listed due to it being an alphabetical index, then they quickly think nothing new is ever there. Or, it's too hard to find. So they're tricking people into thinking there's a bunch of fresh/good content. There's also a cultural phenomenon where everyone discusses "what shows have you been watching lately?" so the Trending aspects of their recommendations is to help people get on board with the trend; and, to push momentum and create the trend too obviously.
> And most people on this planet are more or less happy with whatever they're presented with because they don't care about technology.
I think this is a debatable statement. It could be true, but I am increasingly convinced that enshittification, TikTokification, AIfication, etc. is proceeding despite what the average person wants. Average does not mean gaping, uninspired idiot. I think people in general do notice that everything is broken, short-lived, watered down and ad-ridden. But what to do? When every company does it, voting with your wallet becomes practically impossible.
No, I totally don't mean that people are idiots, I think it's largely ignorance. I, for instance am fully ignorant of audio stuff. I'm mostly happy with Sony/Apple audio products, which audiophiles probably feel the same way I feel about chain restaurants.
It's true that it's also increasingly easier to be presented with an average choice because everything is aggregated somewhere and will mostly converge on a few options.
To your other point, a lot of this is also on an indifference curve. I said what the average person wants, not what the average person is ecstatic about.
But most people don't spend time seeking out the best possible experience and go with the good enough experience they're presented with.
Which is a real problem for the rare person (ie me) who doesn't like McDonalds. Go to a new city and I get recommendations of McDonalds, and the dozen "you won't believe we are not McDonalds" - never mind that I don't like burgers, that is about all I can find when looking for a meal.
True. Though I wouldn't even say it's rare to not like McDonald's. But McDonald's is an option most people are kinda okay with, which is what they optimize for.
Nobody will ever describe McDonald's as a transcendental experience. But it's consistent (same everywhere) and everyone can agree on it (vs. convincing a group to order from a random Indian place).
On HN, we're obsessive weirdos who WILL seek out niche experiences (the interface of this very website is a case in point). But most people aren't.
I just hate so so so much the Netflix of nowadays, they manage to keep me because of a few good movies/series and releasing new seasons of shows that I watched previously.
But otherwise, this interface is so much bat shit! Incredible to me that anyone can pretend to Product manager of something so badly designed and unergonomic.
The most important thing is "continue watching", that should be almost the first line, but no it is randomly spread at different levels. Some times you can't even find it, sometimes it lacks the movie that you were just watching and that reappears later.
It is very hard to find something to watch because they still show you the hundred of things that you saw already, or that old crappy movie that anyone saw ten times on tv, or things that you are not interested anyway.
And there is absolutely no way to filter to not be a frustrating experience.
In addition you have the asshole dark patterns like showing multiple times the same movie/series in a given category when you scroll.
My hypothesis is that they used to have a lot of great content, so that was their strength, and no they have very little valuable and recent content and as they don't want to be upfront about that, they use a lot of dark patterns to confuse you to still give the impression that they have an impressive catalog.
But that has the consequence of the user being frustrated, impossible to find something proper to watch, but still having to spend hours browsing in the app as you might think that the good thing exist but it is just you that can't find it.
It feels to me like they poached some high-level product executive from an intrusive ad company, trained in the art of dark patterns, and pointed them at their paying customers. It's a truly offensive way of looking at your user base, as solely engagement metrics to be optimized. It's what happens when an entire business is built around gamifying one KPI.
I really don't think bad Product Manager's is a good explanation for the UI. Any big company like Netflix is going to heavily A/B test any and every change to the UI. They will only ever add things that boost metrics like engagement. You may not like the UI; it may annoy you, but you should have some appreciation for the fact that they are using sophisticated techniques to optimize for what they care about.
Some people have turned to downloading qBittorrent[0] and use 1337x.to or thepiratebay.org (to start).
At some point these apps are so user hostile that it's simply isn't worth subscribing to. Their margins on content are so low on an individual—effectively zero since a flat fee means ~infinite content—that the effect on their business is incredibly small. Especially for people who have subscribed for months but don't watch consistently.
For movies that are 5+ years old, some would say that the companies have made the vast majority of what they will and copyright is so out of control, bought by those same companies, that it's not bad faith to counter-balance it.
I spent my last 3 months using Amazon Prime on my smart TV, opening the app, scrolling for 15 minutes through the same stuff as last time, turning off the TV and reading a book. I cancelled and now have 15 extra minutes reading time, though I do miss the cheap delivery it got me.
Ironically the old horses were faster! Run XP on modern hardware (if you can get it running at all) and you'll see what I mean. Explorer opens fully rendered in the span of a single frame (0.016 seconds). And XP was very slow and bloated for its time!
It'll do this even in VirtualBox, running about 20x snappier than the native host, which boggles my mind.
It's amazing how fast we can eat up new hardware capabilities. The old 6502 1-MHz CPUs were capable of running much more sophisticated software than most people today imagine, with 1/1000 or 1/millionth the hardware. And now we're asking LLMs to answer math questions, using billions of operations to perform something a single CPU instruction can handle.
The classical answer of why more hardware resources are needed for the same tasks is that the new system allows for way much more flexibility. A problem domain can be thoroughly optimized for a single purpose, but then it can only be used for that purpose alone.
This is quite true for LLMs. They can do basic arithmetic, but they can also read problem statements in many diverse mathematical areas and describe what they're about, or make (right or wrong) suggestions on how they can be solved.
Classic AIs suffered the Frame problem, where some common-sense reasoning depended on facts not stated in the system logic.
Now, LLMs have largely solved the Frame problem. It turns out the solution was to compress large swathes of human knowledge in a way that can be accessed fast, so that the relevant parts of all that knowledge are activated when needed. Of course, this approach to flexibility will need lots of resources.
This is part of why I still have a MacBook2,1 running Snow Leopard. Even with its 4GB of memory and Core2Duo, it's optimized to prioritize my input. It also never changes, which is a form of stability I've come to cherish.
Another point is that you can train a horse, or even eat it if in dire straits. You own that horse. I can't disable things I want to disable, and names, locations, and features change (or are removed) with no notice between minor version updates. I can't tell you the last time I built something for a new Mac, or wanted to.
I don't know MacOS today, and it certainly doesn't make me feel like I own my computer.
I'm less harsh about modern Windows because I view it as amends for Microsoft causing the bot/ransomware crisis of the last 15 years. Still not for me, but at least I neuter it into usefulness.
To be fair even with modern software bloat the overall experience is a lot better now than it was in the XP days. I think it's mainly due to SSDs. They were a huge step change in performance and we fortunately haven't regressed back to the slowness of the HDD era.
At least on most hardware. I have a shitty Dell laptop for work that's basically permanently thermally throttled... :(
Are you running Linux or something? I installed Win 11 in a VM, and no one that's seen its first boot screen would claim the experience has improved since the day of shovelware-bloated XP desktops. It only gets worse and worse from there.
it's ridiculously worse. getting things done on the computer got more tiresome, almost every part (booting is still slow, login is still slow, after login starting the browser is slow, rendering the start menu is slow - which should be in cache, the browser starts fast, but then it takes some time to somehow start loading and rendering the last opened page - which should be in cache, and so on) and yes, you can do a lot more things nowadays "on the computer", but it's mostly "online"
I think they were designed at the time of less powerful machines so they had to be designed better. Nowadays there is not as much push to eke out every last bit of performance because there is loads of power at everyone's disposal and developers are pushed to focus on features first without being given time to refine performance because features mean adoption. So the bloat creeps up, and hardware makers keep designing more powerful machines which further enables the bloatiness. It is a vicious cycle.
You just gave me a hell of an nostalgia hit with "Q6600". Remember when clock speed, cache size and core count were all we cared about? AMD hadn't event bought ATI yet.
Maybe I'll spin up an XP VirtualBox off the back of this thread just for old times' sake and see what happens.
i'm off grid right now and the only fast websites are hacker news, old reddit, and my app https://bookhead.net that is html + a little bit of htmx + a little vanilla javascript
The root problem seems to be monopoly and fragmentation.
When Ford was working on a car, people who wanted a faster horse could go to the horse store. There were reasonable alternatives to Ford's new method of transportation.
But here, you can't recreate Spotify from 2015. You'll never get the rights to play the music for users. Same with Netflix, you'll never get the rights to show the movies.
Same thing with Twitter, Facebook, etc. Even if you know exactly what content your user wants, you can't fetch it for them because it was posted in some other walled garden, and that wall stops you from competing.
If you want a faster horse, change the laws so that people can build faster horses and compete.
Sure, spotify has maintained a near-complete catalogue where Netflix hasnt.
But I no longer find Spotify any good at finding new music, beyond manually looking through artist catalogues.
For context, try out Pandora's recommendations. They haven't improved, yet they're orders of magnitude better than Spotify. The songs are hand annotated for style, content, etc. As a result, they recommend truly new songs with regularity that truly match the vibe.
Compare with Spotify, where everything is based on statistical "people also listened to X". Everything converges on some pop form of whatever genre and songs you've listened to a lot. It'll play odd, out-of context songs from the same artist before it'll find you new artists. Sure they have a few manicured playlists, but its nothing compared to the value Pandora has provided for years.
Spotify only has about half of the music that I like. A lot of it is remixes and jams from Japan that can only be found on youtube, soundcloud, or in some cases it's actually only available on private trackers.
Good luck riding your fast horse through most urban areas (and parking it... er stabling it). All of those things were routine in urban areas before car adoption (I believe Manhattan for instance often had stables in upper floors, leading to some interesting design to get horses up and down).
Readit News
If you're designing a consumer-oriented web service like Netflix or Spotify or Instagram, you will probably add in some user analytics service, and use the insights from that analysis to inform future development. However, that analysis will aggregate its results over all your users, and won't pick out the enthusiasts, who will shape discourse and public opinion about your service. Consequently, your results will be dominated by people who don't really have an opinion, and just take whatever they're given.
Think about web browsers. The first popular browser was Netscape Navigator; then, Internet Explorer came onto the scene. Mozilla Firefox clawed back a fair chunk of market share, and then Google Chrome came along and ate everyone's lunch. In all of these changes, most of the userbase didn't really care what browser they were using: the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family.
So if you develop your product by following your analytics, you'll inevitably converge on something that just shoves content into the faces of an indiscriminating userbase, because that's what the median user of any given service wants. (This isn't to say that most people are tasteless blobs; I think everyone is a connoisseur of something, it's just that for any given individual, that something probably isn't your product.) But who knows - maybe that really is the most profitable way to run a tech business.
Likewise, all social media converges on one model. Strava, which started out a weirder platform for serious athletes, is now is just an infinity scroll with DMs [0]
I do however think that this is an important insight:
> This isn't to say that most people are tasteless blobs; I think everyone is a connoisseur of something, it's just that for any given individual, that something probably isn't your product.
A lot of these companies probably were founded by people who wanted to cater to connoisseurs, but something about the financials of SaaS companies makes scaling to the ad-maximizing format a kind of destiny.
[0] https://www.nytimes.com/2023/12/05/style/strava-messaging.ht...
I mean that's not really the case for paid services without ads like Netflix. They lose money the more you watch. Ideally you'd continue to pay for the subscription but never watch anything.
Non sequitur. For the longest time, Netflix had no advertisements. Do they even now? (I don't subscribe... all their shows end up on my Plex anyway.)
Most great products start out for enthusiasts and often by enthusiasts. They’re opinionated, sharp, sometimes rough, but exciting.
Then VC funding comes in, and the product has to appeal to a broader audience. Things get smoothed out and the metrics rule decisions.
Eventually, the original enthusiasts feel left out. The product’s no longer for them.
So a new product comes out, started again by enthusiasts for enthusiasts. And the cycle repeats - unless someone chooses to grow slowly and sustainably, without raising, and stays focused on the niche.
1. Innovate.
2. Exploit.
You start by innovating a "fast horse". This gains you early adopters who pull in a larger audience. A horse can only be so fast, so continued innovation might lead to something more like a car. This will only cause you to bleed users. Stick to the horse.
Instead of continuing to innovate endlessly, you switch to exploitation. Fire the visionaries. They're just a waste of payroll. Bring in people who can squeeze every last dime out of your user base.
-----------------------
The above isn't anything new. However, it's clear that some companies are better at maintaining quality while exploiting. Are they doing something different, or is it just that their customers have to choose them repeatedly? e.g. Most people don't sign up with one car company for life. They'll buy several cars over their life and that's a choice that the car company must win each time. Meanwhile, people sign up for Netflix or Spotify and stay subbed. They don't look at the alternatives every few years. Porsche needs to keep up with the latest and fastest horses to continue exploiting their reputation, while Netflix can focus purely on making more money from their users. A faster horse may come along, but Netflix doesn't break down and need to be replaced.
If you start the slow growth path at 30 and retire at 65 you will overall make more money from that thing vs someone who sells out at 35. There are some catches though. The person who sells out can go on to the next thing which in sum total may be more sell out enough to make far more over their lifetime, while the slow growth plan you are stuck. The slow growth is over very slow at first, you often spend 10 years making far less than someone who is "working for the man", then 15 more years more or less even, and only then start making good money. There is no guarantee that you will be successful, some people spend their entire life making less than they could "working for the man"; others go bankrupt when a new VC competitor suddenly gets better by enough to take your customers.
There is no right answer. VC money sometimes is the best answer - but many people who reaching for VC money when their better long term answer would be to grow slow.
It hadn't occurred to me that that might actually be exactly what's happening.
The hard bit is to keep taste and discipline at the forefront of design. To not let short-term thinking pollute long-term ambitions. Easier said than done.
> Eventually, the original enthusiasts feel left out. The product’s no longer for them.
I am immediately reminded of when Slack got rid of markdown-style inline formatting, in favor of a WYSIWYG interface, and the internet (or at least, the corner I live in) collectively (and, imo, correctly) lost its shit at them.
There are many reasons for products getting worse; sometimes it's because the company making it is trying to appeal to a broader base, so the original enthusiasts are no longer the target. Sometimes it's because different people are involved and they can't produce to the same level. Sometimes things don't really deteriorate, but new innovations make other things more appealing.
Since more of our culture is in online, advertising-dominated spaces -- the forces of capital have a lot of incentive to ensure smooth growth straight to the sociopath phase.
Maybe the key is just accepting the cycle of it all and ensuring there's always cool new places for creative/enthusiastic people to do their thing.
[1] https://meaningness.com/geeks-mops-sociopaths
Notably, this process usually involves not creating a simplified interface that can be turned into expert mode but actively removing features, adding cues, and steering user behavior though a psychological maze to achieve desired effects.
Basically, people should understand this corporate lifecycle and stop being deluded by the opening moves in a new market that superficially appear to favor customers, individual empowerment, etc. It is a process that always ends in heartbreak because it serves investors, not the public.
My favorite quote ("Marl" is the hypothetical name for the marginal user):
> Marl’s tolerance for user interface complexity is zero. As far as you can tell he only has one working thumb, and the only thing that thumb can do is flick upwards in a repetitive, zombielike scrolling motion.
For street usage, I think those cars are popular because they’re beautiful more than because they’re fast (or because enthusiasts like them).
My utterly soulless Lexus will drive more than fast enough to get me in serious trouble. No one will look at it and feel stirred by its beauty, whereas the typical Ferrari or Porsche coupe will look at least appealing to most and beautiful to many, even those who can’t tell the three marques apart or even unaided recall the name Lamborghini.
Except for the Toyota nerds who will want to come talk to you about the LFA. Ask me how I know!
It’s a mixture of enthusiasm and conspicuous consumption. Most enthusiasts love 90s japanese cars, but the average person sees an old mazda and recoils.
But put an old ferrari in front of anyone and they have a completely different reaction.
OK so we're discussing niche vs mainstream, or "what most people want" vs "what a few want".
The few cars you listed are not popular in the ownership sense, but they are well-known and aspirational.
People can buy them to show off status / money / exclusivity, or perhaps beauty. Speed is table stakes, of course. They have to objectively be better than most cars but also special. They can be strikingly beautiful or strikingly hideous but they must not be ordinary.
If you watch / read reviews of those cars, then it tends to be from the enthusiast driver point of view. Is it good at racing, cornering, reading the driver's intentions and reacting instantly and accurately? But then more often than not, those that can afford them do not buy them to use them for that purpose (or at least not frequently.) Many are treated a bit like investments or merely items in a collection.
What a long-winded way to get back to the original point of faster horses and enshittification of software, eh?
Netflix and Spotify might as well be a Toyota Corolla or Prius. I lost my train of thought. I think I just wanted to pontificate about exotic cars for a while.
(I drive a Polestar 2. It looks like a Volvo, is heavy as a dump truck, but damn is it fast as hell.)
Deleted Comment
You cannot paint by numbers.
By the time our society is collapsing and our rivers are catching fire and our government is being overthrown and our oceans are boiling and our bodies are full of plastic and we can't even escape to another planet because of Kessler syndrome -- all due to their actions -- they'll be old. That will be their kids' problems, and we know the CEO caste fucking hates their own kids.
IE had something like 90% market share back in the day because it was bundled with the OS and cost $0.
Chrome ate everyone's lunch because everyone was using google to search for stuff, and they could advertise their browser on their home page or together with their search results. They also took out ads, in some countries, on billboards, in newspapers and even in cinemas.
I'm sure technical people talking to their families had a small effect (though wouldn't they recommend firefox, because FOSS?), but I think that pales in comparison to google being able to advertise chrome on their search page.
Deleted Comment
That and it was such a better browsing experience. Firefox was not good compared to Chrome for years. I'm sure they are feature parity now, but for years the Chrome experience was significantly better.
They are doing it to increase “engagement” and so more people will stay on their site longer.
Why else wouldn’t Netflix show the “continue watching” row first instead of forcing you to scroll past algorithmic generated crap?
It is the same reason that Google went from describing success as people getting off their site faster and going to one of the “ten blue links” to the shit show it is today.
Presumably the best thing for Netflix is to have a happy userbase, so why do you assume it wouldn't optimize for that?
If I don’t care enough to finish a movie I may as well start a new one. At the very least it’s not a clear choice.
https://nothinghuman.substack.com/p/the-tyranny-of-the-margi...
I might just listen to it, if I have it on in the background, which then in turn feeds the algorithm that it made the "correct choice", but it's a million miles away from, say, listening to a radio DJ where you like their rough output but they're cherry-picking what to play next.
In the world of music discovery a bad recommendation engine is maybe better than a hyper-fine-tuned one.
I have a very horrible case of this. One day at night, I slept listening to lofi playlist. The next week all my recommendations were screwed. Horrible assumption on the part of algorithm.
As long as Netflix was successfully reading the author’s mind, they were satisfied with the experience. However, Netflix assumed that they want to keep watching the same content, oblivious to the author’s desire to discover something entirely new. Netflix failed to meet the expectations of those seeking something entirely different.
I can understand why Netflix made this change. They’ve replaced many shows with their own in-house productions. By doing so, they prevent users from searching for specific shows and then realizing that Netflix doesn’t have them. If this happens frequently, they risk losing customers.
On the other hand, Spotify doesn’t face this issue. Therefore, I’m puzzled by why they’ve made it more challenging to explore content by categories. (Disclaimer: I don’t use Spotify, so my experience is based solely on author’s observations.)
No it wasn't. It was driven by shady crapware distribution schemes and intentionally subtly broken sites under the big G umbrella.
I think this is a great nuance that is often overlooked when discussing this.
Deleted Comment
A very large portion of people actually cares about what they are searching for, and want the ability to ACTUALLY search and find that, with real parameters, not merely get some not-even-close stuff shoved onto their screen instead. That is NOT the serendipity of browsing the stacks in a great library.
A great example of failure is Amazon. I run a small design & manufacturing business, and years ago started getting pestered by Amazon about "Amazon Business" trying to supply both office staples and parts to businesses. This was an area that had enormous potential. Yet, they have entirely failed. I've never bought a single item, and it has faded.
Their primary competitor is McMaster-Carr [0] who does it right. Well-defined categories of everything, and highly specific search capabilities, at reasonable but not bargain prices. EVERYTHING you might search for is fully parameterized in every dimension and feature. Min/max/exact, width/depth/height/thread/diameter/material/containerType/etc./etc./etc. appropriate for each type of product. The key is McMaster DOES NOT WASTE MY TIME. I can go there, quickly find what I want or determine that they don't have it, and get on with my day.
The smaller company that does it right is still beating the tech giant a decade later. Same for other similar suppliers who actually have a clue about what their customers really want.
They continue to prevail over tech giants and VC-funded sites BECAUSE THEY ARE NOT STUPID.
It would be nice if the tech/vc crowd would also stop being stupid. They started out not stupid, but they really lose the plot when they think a few extra eyeballs this week will really win in the long run. At least provide two modes, a strict and serious search and their new messy UI. But they are stupid and this will not happen. Enshittification rules the day.
[0] https://www.mcmaster.com/
The thing that really pissed me off about Amazon Business is that they bought Small Parts and killed it off. Small Parts was a tiny version of McMaster-Carr that specialized in fasteners, small diameter fluid handling, short sections of specialty materials, and in general, quality "small parts."
If I bought directly from Small Parts, I knew I'd get exactly what I wanted. Ordering from Amazon Business? A complete crapshoot. Going to www.smallparts.com now just redirects to an Amazon 404 page!
[/rant]
I mean the answer has to be "they make more money this way" but for me it's means I groan internally before going to Amazon because finding the product I want will be almost impossible - it's even hard if I already visited and already found what I wanted to buy, finding it again, near impossible. Not even basics like search by product manufacturer actually work.
Sites with usable search are a relative joy.
> is the most profitable way to run a tech business.
Yes, I agree. This does seem to be the most profitable model for running a tech business: maximizing user engagement or increasing the time users spend on the platform. Whether that’s achieved through intentionally convoluted UI or by aggressively surfacing certain content, the end goal remains the same.
That said, I don’t think there’s much room left for significant innovation in video streaming interfaces. The core challenge continues to be content — whoever offers the best or most compelling library wins. UI changes might tweak engagement metrics by a few percentage points, but they’re marginal compared to the impact of strong content.
At the end of the day, if there’s a great movie or series to watch, people will show up. If the content isn’t there, no amount of clever interface design will convince someone to spend 30 minutes on something they’re not actually interested in.
> In all of these changes, most of the userbase didn't really care what browser they were using: the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family.
I'm confused as to whether your saying change is caused by catering to the median who doesn't care, or the enthusiast who recommends the latest and greatest. You seem to be saying both.
Except you're making the mistake of thinking these services are optimizing for their userbase. They are not. They are optimizing for revenue and profit growth, a very different target. More ads, cheaper and easier-to-product content, lower opex.
They are converging to churning out the least offensive slop at the cheapest cost with the maximum revenue.
None of the analytics are about what people using the product want, they are about making the most money and growing the fastest. Nothing would look like the services mentioned in the article if they listened to what the users really preferred.
and it matters because this seems to be an omnipresent phenomenon.
everything everywhere seems driven by this unless someone with decision making power is executing a specific and conscious strategy that pushes back against it.
That's the issue, it seems like it really is the most profitable way to do things. Everything sucks now because shooting brainrot and advertisements at our eyes and ears is more profitable than actually giving us what we want.
Indie Hackers is full of people trying to flog their shit AI-powered marketing SaaS, because they've never done anything other than software engineering, so they don't know any good problems to solve. There are uncountably many good problems out there, each with thousands of people who would pay you money to solve them, but those people don't know their problems can be solved by a computer, so you have to go out into the world to find them yourself.
Just like football scouts need to actually visit some niche teams and watch not that interesting stuff to find talent before it is too late.
With tech it might be easier because you might create niche groups so those people come to you.
Just like PG created HN. Nowadays HN is too mainstream so all ideas here are seem already popular so it is like going to scout high school t am that won local championship everyone already knows which players are lined for pro contracts.
Product-Market fit is great if you're developing a SaaS business but it's not necessarily going to give you new inventions — something new is speaking to a potential gap in the market that doesn't currently exist.
Ultimately, analytics are just a view into the business. This thread is complaining about doctors not using microscopes when diagnosing system issues - sometimes a narrow slice is important, sometimes you need to zoom out. If you focus on your "early adopters" or power users exclusively, without understanding how they affect the business, then you are at risk of building things that most of your user base doesn't want.
Power User Analysis: https://andrewchen.com/power-user-curve/
porsche and lambo didn't see the outsized success they have now (financially) until they started pumping out SUVs. hell, the purosangue was made precisely to capitalize on that boring market segment.
i feel there's a little suvivorship bias at play here. i think the important thing is to not forget your enthusiasts perhaps, but a lot of these "successes" wouldn't even be around were it not for appealing to the greater masses. ofc some market segments fare better and you can build a business around enthusiasts.
I think that's a self-dellusion many tech enthusiasts have, that they're somehow trend-setters.
And then the same enthusiasts say for the original iPod "No wireless. Less space than a Nomad. Lame", and see the masses jump to buy it, and themselves only catch up later.
Or they see the masses never caring for their e.g. desktop Linux, whose mass dominance (not mere "works for me" or "have set it up for my elderly parents and they don't even know it's not Windows") would come "any day now" for the last 30 years...
Trend-setters exist, but they're a different group than the "tiny minority" of enthusiasts. More like some musician paid to spot Beats headphones, or some cool dude sporting some gadget early on.
>For example, very few people have any real interest in driving a car at 200 MPH, but Ferraris, Lamborghinis and Porsches are widely understood as desirable cars, because the people who are into cars like those marques.
A hell of a lot of people had a real interest in driving a car at 200 MPH, if they could have the chance. And even more admired Ferraris, Lamborghinis and Porsches because of their design and elegance (and price, people aspire to luxury goods, even when they can't afford them), not because some sport-car afficionados said so.
It's the same in other areas: the popular books, or comics, or movies, or music, etc. are rarely if ever what the "inner" crowd of each niche admires. Most people buy Reacher and such, not Finnegan's Wake.
>So if you develop your product by following your analytics, you'll inevitably converge on something that just shoves content into the faces of an indiscriminating userbase, because that's what the median user of any given service wants.
More likely, if you want to keep and continue increasing your margins, and your stock price, you'll incrementally continue to shit all over your product trying to squeeze ever more money.
Neither the "enthusiasts"/tech-savvy users NOR the "median user" wants Netflix to be the shit it has become, or Google search to be so fucked up, or ads and nags on Windows UI, and so on.
They're just given those, and they accept them having no recourse. The moment there's a better recourse, they jump to it (like IE -> Firefox -> Chrome, or BS early search engines -> Altavista -> Google).
I agree with that.
> but Ferraris, Lamborghinis and Porsches are widely understood as desirable cars
I agree with that too.
> because the people who are into cars like those marques.
I think that is not true. I don’t care about cars. Never had one. Don’t even have a driving licence.
The reason why i think Ferraris, Lamborghinis and Porsches are desirable cars is because they look cool, and they sound cool. They were designed to be like that. If i see one on the street i notice it. I couldn’t care less about the opinion of gearheads. If a car would come out looking like my grandpa’s skoda, but all the car lovers would love it I wouldn’t even hear about it.
It is all about flashyness of the industrial design. And rarity of course.
I think the big difference is that nobody is going to pay $10m for a web service or browser.
Once you become a bloated monopolist like the three companies you just mentioned, your distribution strategy is solved in other ways (like, you've done some bundling and some acquisitions, maybe pressured a few companies into exclusivity agreements and are probably breaking some anti-trust law or other but you have lawyers). Then you don't care about the experts, PR or niches anymore, and you serve up slop. When the analytics recommend slop you go with the analytics, when they don't you ignore them.
None of this is to discount your insightful comment, just saying once you're big enough, your strategy is just doing tricky distribution deals, really (a fact no record executive would dispute).
It's probably profitable in a lot of cases to follow those metrics, shovelware content is cheaper to produce, and since the median user pays the same subscription fee as the enthusiast, you get better margins producing slop for the uncaring masses.
You need enthusiast businesses owners to produce quality product.
Damn, I never thought of this before, but it explains so much!
I promise it is NOT always a good idea to follow the enthusiasts, because they are not at all like everyone else who uses your thing. Following them will skew your decisions—unless they are your entire customer base, so, have at it.
This article imo is complaining about the effect of middle management product owners at large companies. There are two dynamics that both converge on enshittification:
1. These product managers (or product designers) are early in their careers and want to make a splash. They are given lower priority projects but try to break out by making them bigger, better, more non-horse-like. They over-design and over-complicate the solutions as a result, because they don’t yet know when the right solution is just a refinement of what’s tried and true. They are incentivized to do this because they want to break out of the mold.
2. The managers above them, or a layer or two above depending on company size, are risk AVERSE. They are tasked with delivering results regularly and consistently. If you have the innovation bug or are creative at this layer, you get moved onto projects where this is required, which is not most of them. Overcomplicated is fine sometimes with you but WEIRD is absolutely not okay (the stuff that actually could be innovative), and no one gets fired for following The Metrics.
These two incentives clash to create overcomplicated but functionally poor products that aren’t helping anybody out. A healthy skepticism of complication and a healthy skepticism of engagement as the sole metric (or metrics in general) is necessary to make good shit. Sometimes it is actually understanding and using things as an enthusiast would, but you need to bring in an understanding of how the rest of your users are distinctly different from the enthusiasts, too. Using your thing yourself and actually following your own subtler feelings is what produces really useful innovation, slowly and surely over time.
The Nash Equilibrium of streaming UIs is a TikTok experience.
:-(
There are systems out there that can do AB/CD testing and those do a better job of finding pairs of changed that have compounding effects.
You cannot A/B test your way from chocolate and peanut butter to cherry and vanilla. So we get to deal with tone deaf companies who feel their analytics are proving that customers either don’t know what they want or are lying about what they want. But that’s not something A/B testing can prove. It takes more sophisticated experiments than that.
... primarily for their price tag. There are a lot of enthusiasts for money in the world, much more than for driving at 200 mph.
> the change was driven by enthusiasts recommending the latest and greatest to their less-technically-inclined friends and family
It was never about recommendations. MSIE and Chrome were (and are, but with Edge Browser instead of MSIE) shoved into consumers' throats by ads, marketing, bundled distribution and outrageous lies.
Something is popular, folks are envious of it, they end up building something much like it. Doesn’t matter if it’s houses, logos, or user experiences – seems to be how things work.
1. I wouldn’t say the car veands you mentioned are popular because they can hit high speeds. In my experience nearly any car can with the right engine and equipment in it (of course due to weight distribution and other details I assume they’re not all equally safe but that aside).
Personally when I look at those brands I think they’re sleek and pretty and when I feel like wanting one it’s because they’re expensive cars, driven by the rich. They’re not chosen only by the rich cause they have the best taste, they’re chosen by the rich because they are the only ones to have the financial means to afford one.
Also I feel like the changes made based on analytics arent made to please (more) users but to make as much money as possible, whether that be pleasing users in the starting phases of your company or in the latter phases when you already dominate the market squeezing money out of your big existing userbase.
This is so spot on. I was a long-time serial entrepreneur who spent a couple decades across three successful startups discovering, shipping and growing new categories of tech products primarily for consumer, prosumer and hobbyists. Then I sold my last startup to a very large F500 silicon valley tech leader and ended up a senior product exec there. While there were a lot of positives like more mature engineering processes, testing and devops as a discipline, the exact issue you describe was a nightmare of product-damaging mistakes I called "analytics abuse." In my startups I valued having increasingly robust analytics over the years. In part because they helped increase my overall understanding of usage but mostly because they provoked good questions to explore. That exploration happened naturally because as the "product guy / founder" I never stopped spending a lot of time with our most passionate, opinionated, thought-leading customers. Over years of iteration I'd learned how to engage deeply and listen carefully to input from these customers. This involved interpreting, filtering and curating the mess of divergent personal preferences and pet feature ideas to tease out the more actionable product signals that could increase broad usage, adoption and passion around our products. I'd then bring those curated signals back to the product teams for evaluation and prioritization.
At BigCo they were diligent about meeting with customers, in fact they had entire processes around it, but their rigorous structures and meeting agendas often got in the way of just directly engaging and actively listening. Worse, the customer meetings the more senior product decision makers actually attended in person were mostly with the highest revenue customers. Junior PMs (and sometimes new grads) were delegated to meeting with the broader base of customers and filing reports. Those reports were then aggregated by ever-helpful program managers into tables of data and, eventually, slides - losing all nuance and any ability to spot an emerging outlier signal and tug on that thread to see where it goes.
I tried to convince everyone that we were missing important customer signals, especially from our smartest, most committed users. Being only one level removed from the CEO and quite credible based on prior success, I was definitely heard and most people agreed there was something being lost but no one could suggest a way to modify what we were doing that could scale across dozens of major products and hundreds of product managers, designers, execs and other stakeholders. In my experience, this general problem is why large companies, even the most well-run, successful ones full of smart people trying their best, end up gradually nerfing the deeper appeal in their own products. Frustratingly, almost every small, single step in that long slide pushes some short-term metric upward but the cumulative effect is the product loses another tiny piece of the soul that made our most evangelistic, thought-leading customers love the product and promote it widely. Ultimately, I ended up constantly arguing we should forego the uplift from some small, easy-to-prove, metric-chasing change to preserve some cumulative whole most people in the org weren't fully convinced even existed. It was exhausting. And there's no fighting the tide of people incentivized on narrow KPIs come bonus season.
I'm sorry to report I never found a solution to this problem, despite my best efforts over several years. I think it's just fundamental. Eventually I just told friends, "It's a genetic problem that's, sadly, endemic to the breed" (the 'breed' being well-run, very large tech companies with the smartest product people HR can hire at sufficient scale). Even if I was anointed CEO, given the size of the product matrix, I could only have personally driven a handful of products. I do think codifying premises and principles from the CEO level can help but it still gets diluted as the number of products, people and processes scales.
I'm groping towards something vaguely ombudsman-y, or WW2 production/logistics trouble shooters. Or maybe even pre-Bush41 ARPA Project Managers - term-limited person-with-a-checkbook and few accountability constraints.
If one accepts this role has to be out-of-band, vs poking big hairy blob in hope of creating and maintaining signal channels with particular properties, and grants CEO-adjacent leverage, then it seems a remaining unresolved challenge is integrating the output signals at scale? If so, maybe (jest) CA granted KPI offsets?
That risk (losing all content and facing extinction) is what pushed Netflix in the direction of being a content-producer, rather than a content aggregator. I agree with everyone's points on the influence of the median user in diluting the quality of the content Netflix produces, but that's not the only forced that pushed us here. Spotify faced a similar crossroads and decided to broaden beyond music once they started losing bidding wars for licensing.
Being a faster horse wasn't an option available to either Netflix or Spotify; there is no path for a 'better 2012 version of netflix or spotify' in 2025. They each had to change species or die, and they chose to keep living.
I wasn't aware that Spotify lacked much in the way of mainstream western music.
Are they having licensing issues?
Did they, though? 2025 Netflix is extremely close to having a worse UX than piracy, and it's already far more expensive. Are people going to pay a fortune for Netflix when their handy nephew can hook them up to his far superior Jellyfin instance for a sixpack of beer?
It's a tragedy of the commons, really. The whole value is in having a complete catalogue available for the casual viewer, and making $10-$20 from someone wanting to watch a random decade-old movie twice a month or so. Break up that catalogue into twenty different services each charging $15, and that same casual viewer isn't going to subscribe to a single one of them.
If the streaming industry doesn't get its shit together they are either going to lose viewers to piracy, or to a completely different medium.
Netflix I understand much less. They make money from subscriptions. If you perceive having a fantastic experience on the site by just going there, finding something you enjoy watching, and leaving... they win. Why they would foster a doom-scrolling experience I really can't really explain, other than imagining some dark pattern like they have to pay per view and want you to watch C grade movies? More time spent looking for something to watch means less time streaming?
I don't get it.
You can't provide a seamless UX for turning on the TV and watching The Office if you don't own the rights to The Office. They want to habituate you to scrolling through content Netflix actually owns and picking something, because it's apocalyptic for them if you ever treat the services as fungible content libraries that you hop between month-to-month.
A short while ago, I noticed I only used Netflix to watch 2 classic comfort shows, and I started to doubt if it was worth a 2-classic-comfort-shows-as-a-service subscription. I tried looking through the catalog to see what else I was paying for and ended up cancelling my subscription.
Netflix does an amazing job in giving the impression that they have an endless library of top quality content, but in reality, it seems like it's only a handful good shows and some filler, but presented in a way that makes it look like there's way more than it actually is.
or, if you're presented with more random 'clips' or movie snippets, this turns on your gambling reward center. It's like a slot machine - where you "win" by finding a good series to watch after searching. And because this is random, you end up getting addicted to looking thru the list/snippet, trying to encounter a perfect series to watch.
They want to take the bargaining power from creators (and old IP owners).
They don't want the customers to search for a specific show. They want the customers to watch whatever is shown to them. This way Netflix will have tremendous power over show creators - if our algorithm doesn't favor you, it doesn't matter how good your show is or how much money you spend on marketing outside Netflix.
Attention destroying apps reduce the long term focus and reward centers such that doom-scrolling through the catalog probably feels better than just watching something. Most of the folks I know who start a movie or show immediately pull out their phones anyway to scroll elsewhere.
Because my netflix subscription is cancelled specifically because the "Finding something I want to watch drains my energy" phenomenon. Gradually over the course of like a year I got more and more frustrated with being suggested things, and not having a good way to find things.
Apple TV is the worst, because it dumps you right into the program, and you have to back out in order to get more information.
They all just want me to trust them that I'll love it. I end up having to pull up reviews on my phone.
That's depends on your definition of "want". They might not want to on, but their monkey brain does.
Because regardless of whether or not the business model depends upon it, investors have been trained that “engagement” is inherently good quality for their investments to have. Increase engagement, stonk price go up.
https://www.macrotrends.net/stocks/charts/NFLX/netflix/net-i...
Maybe it is winning despite what Netflix leaders are choosing to do, and maybe their choices will cause them to falter soon. And maybe Netflix could be doing better than they are. But it is always easier to pontificate than execute.
I don’t buy Netflix solely because they don’t integrate with the search in the iOS/macOS TV app.
Unfortunately, based on media trends before streaming and Netflix was a thing, lots of people like C grade productions. If you recall, “reality” TV shows were taking over in the 2000s. People like the Tiktok-ificiation (or otherwise lowering of quality).
I quit all those platforms recently and I'm not missing the frustration of having to 'switch channels' through their incomprehensible categories and views anymore.
Many people will pay Netflix for years hardly watching content for months just because the convenience factor of not having to subscribe/unsubscribe when they know a new season of X will be out in the next year. It's wild to me, but people are lazy. So again, the more you keep them from actually watching the content and realizing they are "done", the longer they likely just keep their subscription active. Get them to add as much potential content they want to watch to a never ending backlog watch list.
Of course, such a feed would take me 2 minutes per week to read through so that wouldn't be good for the business.
Entertainment is a zero-sum market. More time spent doom scrolling means less time spent on another service, which probably reduces their churn (also, ads)
i get it, i hate what they've become too. i'd like to believe there's a world where paying for content is a better model than selling ads. but the reality is that every time netflix makes a decision that the internet gets angry about, their balance sheet looks better.
You can see how Sonos enshitified their interface and even with a user rebellion wouldn't back down, just as an example.
To keep the line going up, platforms have to appeal to wider and wider swaths of a population, eventually lapping at the shores of a population that really doesn't care or want this service. But if you can hook them with some dopamine in a 5-second video, or a quest to rediscover some neat thing that they saw two page-loads ago but is now mysteriously gone from the very same list it appeared in, then you've clawed one additional user into your metrics and the VCs give you a treat.
These people don't care about the service and they're the worst users to cater to, but everyone caters to them because they're the only ones left. Hence, TikTokization.
What I don't get is at some point the marginal user increase for a change has got to be smaller than the number of customers you tick off and lose by changing things.
Is the idea that all services converge on the same N billion people target audience who wants something almost entirely unlike the initial product? I feel like "marginal" doesn't really capture this nuance if so.
https://nothinghuman.substack.com/p/the-tyranny-of-the-margi...
I finally know what to call these idiotic trends that I've learned to recognise but couldn't name.
The one that grind my gears the most has been Microsoft breaking decades-old Windows paradigms to cater for Linux-developers-on-Windows, which is a very marginal, even actively disinterested group. All this at the expense of the 99.9% of their loyal user base.
For example, VS Code had the opposite shortcut (literally with the arrow keys going in opposing directions) for "go back in search history" to every other editor ever made for Windows... but matching the Linux equivalent.
Similarly, they recently broke "cls" to match the broken(!) behaviour of "clear" in Linux because of basically just one or two Linux users complaining in a GitHub Issue ticket. Windows users weren't listened to, because they're already users, not potential new users.
[1] oh, what the hell!? pun unintended, but now at least the name makes sense
Consumer apps at massive scale like TikTok and Netflix don't design for nerds like us, they design for the average person. Actually, they design for the average behavior of the average person.
And most people on this planet are more or less happy with whatever they're presented with because they don't care about technology.
And when you control what's presented to people, not they (and they don't care), you can push them to consume what you want them to consume.
I heard a YC group partner once that he's worked with a ton of delivery apps. Many of them start out as differentiated apps for ordering from the best "hole in the wall" places or the app for authentic foreign cuisines, only to discover that the best growth hack is getting McDonald's on the app, because that'll be your top seller, instantly.
Most people just do the default thing everyone does—and we're probably all like that in one aspect or another of our lives, and that's who many experiences are designed for.
But engagement maximization looks the same everywhere – it’s communicating with the amygdala of the user, not their consciousness. And in a way, everyone’s amygdala is kind of the same and generic (sugar foods, violence, rage bait, boobs, chock value etc). Products that are largely designed for higher consciousness are more varied, such as most books. But those drive less engagement.
The amygdala wants more of the same, and the prefrontal cortex seems to want variation. My view is that you can’t have the chocolate muffins and raw carrots on the same plate, or a bookshelf with both Dostoevsky and Playboy magazines. You have to compartmentalize to protect yourself from your own amygdala. Same goes for media. Even well meaning product managers will be completely fooled if they simply follow the metrics.
And part of the problem is that if somebody (TikTok) has the most engaging format possible (vertical short-form video) and you (Substack, Reddit, LinkedIn, etc.) don't, you're at a strict disadvantage. So you enable short-form video, boost it in the algorithm, etc. no matter if it's a fit with your product because people will watch it if it's put in front of them.
> My view is that you can’t have the chocolate muffins and raw carrots on the same plate, or a bookshelf with both Dostoevsky and Playboy magazines.
And the problem is that in media, the prefrontal cortex stuff will never make as much money as the amygdala stuff, so few platforms will survive by focusing on the prefrontal cortex stuff.
A big reason HN is still so cozy and surfaces cool articles and discussions is because YC doesn't have to monetize it or optimize for engagement.
But imagine trying to start HN today...
One can always do as in "The good place" show: put a bunch of hotties to talk about and play with moral philosophy. I think the show was somewhat evil in that approach, but at the same time, it was also morally sound...
At least McDonald's doesn't pretend.
They're generally designed for engagement. Nobody is particularly asking for this type of experience it's just that Tiktok has discovered the most addictive - eh hum, I mean engaging - experience thus far. So they're being copied.
Netflix is a little different though as if people open the app and always see the same top titles listed due to it being an alphabetical index, then they quickly think nothing new is ever there. Or, it's too hard to find. So they're tricking people into thinking there's a bunch of fresh/good content. There's also a cultural phenomenon where everyone discusses "what shows have you been watching lately?" so the Trending aspects of their recommendations is to help people get on board with the trend; and, to push momentum and create the trend too obviously.
I think this is a debatable statement. It could be true, but I am increasingly convinced that enshittification, TikTokification, AIfication, etc. is proceeding despite what the average person wants. Average does not mean gaping, uninspired idiot. I think people in general do notice that everything is broken, short-lived, watered down and ad-ridden. But what to do? When every company does it, voting with your wallet becomes practically impossible.
It's true that it's also increasingly easier to be presented with an average choice because everything is aggregated somewhere and will mostly converge on a few options.
To your other point, a lot of this is also on an indifference curve. I said what the average person wants, not what the average person is ecstatic about.
But most people don't spend time seeking out the best possible experience and go with the good enough experience they're presented with.
Nobody will ever describe McDonald's as a transcendental experience. But it's consistent (same everywhere) and everyone can agree on it (vs. convincing a group to order from a random Indian place).
On HN, we're obsessive weirdos who WILL seek out niche experiences (the interface of this very website is a case in point). But most people aren't.
In categories where neither is the case, you can usually find beautiful alternatives from indie makers or small businesses.
The issue with streaming and social media is that they represent 90%+ of our cultural narrative now, so it feels like there's no escape.
But otherwise, this interface is so much bat shit! Incredible to me that anyone can pretend to Product manager of something so badly designed and unergonomic.
The most important thing is "continue watching", that should be almost the first line, but no it is randomly spread at different levels. Some times you can't even find it, sometimes it lacks the movie that you were just watching and that reappears later.
It is very hard to find something to watch because they still show you the hundred of things that you saw already, or that old crappy movie that anyone saw ten times on tv, or things that you are not interested anyway.
And there is absolutely no way to filter to not be a frustrating experience.
In addition you have the asshole dark patterns like showing multiple times the same movie/series in a given category when you scroll.
My hypothesis is that they used to have a lot of great content, so that was their strength, and no they have very little valuable and recent content and as they don't want to be upfront about that, they use a lot of dark patterns to confuse you to still give the impression that they have an impressive catalog.
But that has the consequence of the user being frustrated, impossible to find something proper to watch, but still having to spend hours browsing in the app as you might think that the good thing exist but it is just you that can't find it.
The cost of maximizing "value" for the company to the nth degree degrades the customer experience once it exceeds a certain threshold.
It's greedy tunnel vision that makes the world worse for everyone in the long term.
At some point these apps are so user hostile that it's simply isn't worth subscribing to. Their margins on content are so low on an individual—effectively zero since a flat fee means ~infinite content—that the effect on their business is incredibly small. Especially for people who have subscribed for months but don't watch consistently.
For movies that are 5+ years old, some would say that the companies have made the vast majority of what they will and copyright is so out of control, bought by those same companies, that it's not bad faith to counter-balance it.
Not sure. These are arguments.
[0]: https://www.qbittorrent.org/
Sooo much better.
This seems to be common among the streaming services. I can't imagine any reason other than they want to force people to see their other content.
It'll do this even in VirtualBox, running about 20x snappier than the native host, which boggles my mind.
This is quite true for LLMs. They can do basic arithmetic, but they can also read problem statements in many diverse mathematical areas and describe what they're about, or make (right or wrong) suggestions on how they can be solved.
Classic AIs suffered the Frame problem, where some common-sense reasoning depended on facts not stated in the system logic.
Now, LLMs have largely solved the Frame problem. It turns out the solution was to compress large swathes of human knowledge in a way that can be accessed fast, so that the relevant parts of all that knowledge are activated when needed. Of course, this approach to flexibility will need lots of resources.
Another point is that you can train a horse, or even eat it if in dire straits. You own that horse. I can't disable things I want to disable, and names, locations, and features change (or are removed) with no notice between minor version updates. I can't tell you the last time I built something for a new Mac, or wanted to.
I don't know MacOS today, and it certainly doesn't make me feel like I own my computer.
I'm less harsh about modern Windows because I view it as amends for Microsoft causing the bot/ransomware crisis of the last 15 years. Still not for me, but at least I neuter it into usefulness.
At least on most hardware. I have a shitty Dell laptop for work that's basically permanently thermally throttled... :(
I use virtualisation for the rest.
Maybe I'll spin up an XP VirtualBox off the back of this thread just for old times' sake and see what happens.
When Ford was working on a car, people who wanted a faster horse could go to the horse store. There were reasonable alternatives to Ford's new method of transportation.
But here, you can't recreate Spotify from 2015. You'll never get the rights to play the music for users. Same with Netflix, you'll never get the rights to show the movies.
Same thing with Twitter, Facebook, etc. Even if you know exactly what content your user wants, you can't fetch it for them because it was posted in some other walled garden, and that wall stops you from competing.
If you want a faster horse, change the laws so that people can build faster horses and compete.
Spotify has almost anything I look for. Netflix I struggle to find anything of interest.
But I no longer find Spotify any good at finding new music, beyond manually looking through artist catalogues.
For context, try out Pandora's recommendations. They haven't improved, yet they're orders of magnitude better than Spotify. The songs are hand annotated for style, content, etc. As a result, they recommend truly new songs with regularity that truly match the vibe.
Compare with Spotify, where everything is based on statistical "people also listened to X". Everything converges on some pop form of whatever genre and songs you've listened to a lot. It'll play odd, out-of context songs from the same artist before it'll find you new artists. Sure they have a few manicured playlists, but its nothing compared to the value Pandora has provided for years.
Dead Comment