I wonder what is going on at Strava. It feels like peak Strava was some time around 2018. Since then, it seems like features have only been getting disabled, hidden behind opt in flags, or moved to the paid plans. I completely understand the decision to move things to paid since their service does cost a lot to run and the subscription isn't expensive, but where are all the new and cool features that should have come out over the last few years?
And I wonder if the article is correct in that they are pivoting away from segments. That seems strange since segments seem like the most compelling feature.
Strava is extremely frustrating. I've been a paid user for 10+ years. They do add features from time to time, but it's amazing how slowly it happens. There's a forum on their web site for user feedback and suggestions, but it seems like they pay little attention to the users. So many of their users that I know feel like a run or bike ride didn't happen if they don't record it on Strava. It's a super addictive and sticky product, yet Strava seems to squander this loyalty by ignoring their customers.
Route building and personal heatmaps are some of my favorite features, but there's so much that could be added (like a simple search). To some degree their customers are spread out over so many market segments beyond the original biking and running that they started with. Each segment has their own needs and wants. Then there's all the hardware (watches, power meters, trainers, etc..) that needs to be integrated and supported.
A few years back, I noticed a morbid topic in their forums about a user requesting a feature but a few years later getting diagnosed with a terminal illness before Strava delivering on the feature[0].
I posted it on HN but got flagged[1]. I still find it eye opening, a reminder that the code we write are products impacting the finite life of other people. Maybe we should have more modular, more open software as a standard so that other people can work on and deliver experiences that the original creators fail to deliver.
I'm an occasional strava user because I wanted just something to log these data and strava seemed like the popular choice, but what a clunky app. First off there are so many features and I have no clue how to use any of them, half of them ask for a subscription. Feels like my junky cut rate wyzecam app always asking to upcharge me, or maybe Spirit Airlines. The auto pause feature rarely works correctly (maybe not at all?) on my cellphone. I'm sitting there still as a statue and I'm watching average speed tick down...
I used it snowboarding and there's some way to track runs and push out the whole day of skiing as one post, but I couldn't figure it out, so I just polluted my own feed with these single run posts. Then for some of them I forget to start recording halfway down the run because once again thanks to the broken auto pause feature, so now I need to remember to start it up at the top of the hill in addition to already having to put on a snowboard binding. Then thanks to the auto pause and me forgetting about pausing the run, it records me shuffling about in the lift line and then riding 25mph on the nose on the gondola.
Most apps/startups begin as an exploration of product/market fit. They'll try a lot of different things and use analytics (even simple server-side stats) to determine what people actually use in the app. Very frequently, you discover that the things you thought your users would want are actually only used by 0.05% of your customers. Eventually you have to start shedding rarely used features and limiting free plans, even if it makes the non-paying users angry.
Truth is, it doesn't really matter if you're losing someone who spend 5 years on the free plan but refused to sign up for the paid plan. They're not converting to paid unless they're forced to, and you're not gaining any money by letting them stay on the free plan for another 5 years.
> but where are all the new and cool features that should have come out over the last few years?
Cutting rarely-used features like this one could be a sign that they're trying to free up engineering resources to ship new features.
> They're not converting to paid unless they're forced to, and you're not gaining any money by letting them stay on the free plan for another 5 years.
This is true, but it ignores the importance of network effects in products based on social media/interaction. I'd guess that a huge part of the attraction and user retention ability of Strava is the social aspect, which they'd be crippling if free users migrated somewhere else. As a secondary issue, Strava also benefits from the data generated by free users, though I have no idea what the value of that data might be.
From discussions I've had with people at Strava, they're trying to be more focused on implementing a few key features really well. When the original founders returned in 2018 or 2019, the app had a ton of features that had been half integrated, couldn't be found in order to be used, etc etc.
They paired back things that weren't working, moved to focus on revenue (as they weren't going to be able to grow into the social network for all activities like they tried).
I think it's the right strategy, but this move makes no sense. It's very easy for them to manage duplicate data, and I'm surprised to see them suggest users set-up their garmin device to connect directly to Apple. This essentially makes it easier for strava to be cut out of the loop in the future if the social aspect isn't working for you.
Re: "key features really well", they're not even doing this.
My normal morning jog goes under a large bridge. This often confuses my phone's GPS, leading it to return a spurious point or two a kilometer away. Strava is completely incapable of detecting these spikes (accelerating to 200 km/h when running is normal right?) and offers no way to edit them out afterwards, meaning all my speed and distance records, progression charts etc are ruined by junk data.
> And I wonder if the article is correct in that they are pivoting away from segments
That's not what DCR wrote, that part was clearly about user perception: Strava has grown into something bigger than just that app that declares you KOM when you go really fast.
There's also the reality where their client base gets older and loses interest in taking their sport as seriously. My entire cycling circle stopped regularly using Strava years ago because we just all got older and no longer had that competitive interest.
Started using it in 2019 and it only got worse, it's a shame because I really enjoy the social aspect of it since it bundles a lot of sport and gear into one place but with this trend I feel like the users will be leaving.
All of the (Garmin, Strava, Apple, and Google [who is trying really hard but failing]) want a monopoly on your health data. Garmin is the only one I actually trust currently, as they've shown 0 willingness to monetize your health information.
The Strava/Garmin relationship is an interesting one: Strava has the social network, Garmin has the best devices [for serious Athletes, not casual users]. Garmin Connect is pretty cool in it's own right, but the Social features never really took off, which is where Strava plays and has a de-fact monopoly. Strava can't survive without Garmin, and Garmin benefits from Strava's content.
Garmin has pulled some 'power moves' in the past though with people it doesn't like... A competitor: Wahoo, who made cycling GPS computers, was cut off from inserting data into Garmin Connect and it left a lot of users out in the cold. Most serious cyclists will use Garmin devices, not an Apple watch, to track their rides as it seamlessly integrates with ANT+ sensors: power, cadence, wheel speed, heart rate, chainring and cog positions sensors.
Interesting to see Strava cut off Apple... I'm guessing it has to do something with preventing them from developing an alternative to the Strava social network.
Strava isn't cut off from Apple. Instead, Garmin/Fitbit etc. would send their data to Strava, which then would forward it to Apple for them. The part where Strava takes the data from non-Strava apps is the one that got disabled (presumably by strava and not by garmin, but either could have).
Garmin couldn't directly disable this, once Strava has your data they can do with it whatever they want, licensing permitting. Which is a segue into - while I have no specific knowledge of this case, I do think it's somewhat tricky to take data from a third party and then hand it on to another third party. Presumably all of this was mediated by the user, and user's desires should be paramount here, but it's an example of why I believe OAuth is fairly broken and what we should all be giving permission to is capabilities, not access.
Really surprising to see Strava actively avoiding the central switchboard role that would be such a powerful retention feature. I wouldn't be surprised to see it come back very soon, with some low level UX polisher receiving that talk about power and responsibility.
> All of the (Garmin, Strava, Apple, and Google [who is trying really hard but failing]) want a monopoly on your health data. Garmin is the only one I actually trust currently, as they've shown 0 willingness to monetize your health information.
I guess their huge data breach that they didn't notify and spent weeks fixing doesn't phase you, but it punted me out of their system.
I'm still shocked and amazed that anyone would willingly share their health data with corporations. Why? What could go right? I say that acknowledging your point that this is possibly one of the better corporations out there.
> Garmin is the only one I actually trust currently, as they've shown 0 willingness to monetize your health information.
What about the concern that the company is bought out in the future? Or that it may be sharing data already, with government agencies, etc? Or companies that it works with?
I honestly can't even imagine the criteria whereby a corporation could be 'trusted' with personal information!
It boggles my mind, that people think about this - and plainly you do - but come to the judgement that its ok for corporations to have this personal data!
>> Garmin is the only one I actually trust currently, as they've shown 0 willingness to monetize your health information.
> What about the concern that the company is bought out in the future? Or that it may be sharing data already, with government agencies, etc? Or companies that it works with?
> It boggles my mind, that people ... come to the judgement that its ok for corporations to have this personal data!
I'm not OP. I chose Garmin precisely because I don't think it's ok for corporations to have this personal data.
I chose Garmin because I can use it without needing to share any* of my data with a third party, including Garmin itself. I save the workout files to my computer via the same USB cable I use to charge the watch. There are various non-cloud apps I can use to view and analyze these workouts, if I actually cared to do so beyond the "fastest 5k" etc that the watch tracks automatically.
For this, it does not matter to me whether Garmin is bought out in the future because my watch works just fine as-is, and cannot update its OS without my explicit permission. I'm unsure what data Garmin would be capable of sharing because I have given it none.
* of course, speed/time/location data is obtained by pinging GPS/GLONASS satellites, but the watch can and does record my workouts quite accurately in non-GPS pedometer mode.
I prefer the corporations not have the data at all. The problem is Apple is simultaneously the best and also pretty bad at stewarding such an ecosystem. In theory all my actual health data would live on a device I own, encrypted, and accessible only by the apps I approve. In reality Apple can't or won't provide those mechanisms. Every other corporation wants to collect and sell my data.
Apple had a chance to make on-premise data storage a thing. They had AirPort, which combined with a Mac Mini could provide full iCloud functionality from the home. iCloud itself could have simply been an encrypted offsite backup, like Tarsnap with no options.
Instead Apple built a system that still puts clear user data within their reach.
As long as they don't reveal my 'hidden zone', they're welcome to the lot, and I freely give it whilst severely understanding the 'risks'. If my W/kg or bpm can be monetized, go for it.
Hit me with one single ad, though, and it's goodnight Vienna.
Many people don't care about health data being monetized. I don't personally care if Strava sells my heart rate data or where I rode my bike today. I'm fine with trading them that data for their heat map, which is super useful in planning rides.
To me, it comes down to me being fine with this data being considered public. If others don't want to share it, that's totally fine and I understand their perspective. For me though? I couldn't care less.
Congrats, you sold some data that says I'm fat, out of shape, and don't like riding my bike on busy roads.
I think this can make sense. The recording devices ideally should have their own integrations with Apple Health so the source pushes to all the data repos rather than repos chaining the data along.
Wonder if this includes third party sensor data that’s tracked with Strava. Like if I use a chest strap and record into Strava, will the exported data include the sensor data or is that considered “third party”
The weird thing is that they claim it's to avoid duplicate activities, but they totally know how to recognize duplicates already. Every so often, there's some sort of glitch in one of my activities getting from Garmin to Strava. If I'm feeling impatient, I download the .fit file from Garmin and upload it to Strava myself, and I never get a duplicate that way. Happened for the first time in a while just last week. Clearly, whenever Garmin does send the data, Strava is perfectly capable of recognizing an activity it already has, and it does the right thing. I'm just not buying that excuse.
BTW and a bit OT, I find it very impressive that Strava can retroactively create a leaderboard going back years for a newly created segment, meaning that they must evaluate potentially millions of nearby activities for overlaps, often in just a few minutes. That's a hell of a query. Anybody know of more information on how they do it?
It's not the size (though as mdoms explains it's more than you might think); it's the complexity of matching activities to segments at the necessary level of precision. I know what a consequential amount of data looks like, having worked on a storage system that routinely fed multiple petabytes of data to each of many analytics pipelines simultaneously. Geographic partitioning is only what gets you to millions of activities instead of billions. There's still a computational component involved in delivering those answers so quickly, and that's what I'm curious about. There might be some interesting algorithms involved. Your blithe dismissal of any part other than the one you think you understand shows that you either can't understand the whole thing or didn't bother to try.
Incredibly ignorant statement. An activity can have thousands of data points across dozens of metrics. A record of your route to 1 meter or smaller precision for a 5 hour bike ride is alone thousands of data points. Along with hundreds or thousands of time series points for heart rate, elevation, cadence, power, respiration, etc etc... Millions of activities means tracking tens of billions of data points and perhaps an order of magnitude more than that.
I thought they meant duplicate activities on the Apple Health side. If they can't read what the other services are writing to Apple Health then they have no way of avoiding duplicates.
Doesn't that suggest that Apple is failing to detect and coalesce duplicates, like Strava and (AFAIK) other similar services like Garmin Connect or old MapMyRun do? That seems pretty damning, and also not Strava's problem.
Instead of fixing shit like the issues pointed out in this post, they spent all their resources fucking up the UX in the app. The new activity save screen is complicated and is very obscure about where the various form data are going to show up in the final post. It used to be intuitive and straight forward. Feels like they were trying to be clever.
What is the most important data (or strava features) you want to keep?
I was actually looking into this myself, and a few parts of strava are not that hard.
For me the overview of my rides (including avg speed and distance is important), I quickly implemented a webpage with javascript and leaflet (map viewer), and I could present that data pretty quickly using geojson. I had an offline program which converted gpx traces to geojson, but I'm pretty sure I'm able to read gpx in javascript directly.
The harder part is generating the gpx trace (which I used komoot app for), as this involves matching your trace (gps positions) on a map instead of using the data as is.
I actually spent a lot of time working on one. As usual there isn't a whole lot of special sauce that makes it impossible but its an immense amount to replicate all of the features as well as a rather expensive to run piece of software if you want something like segments and high score boards.
And after all this, your user base consists of the people who didn't want to pay for strava.
I built one for myself, sqlite + apache superset. I basically did that because over the years I had switched devices from different companies and I used to use endomondo which was shut down. I wanted to have all my data in once place. It does require me to export data from device and use my import script once a while, however works for me.
Progress is slow as a solo dev who contributes to it in their spare time, but I’m happy enough with it that I could archive my Strava account and switch to it fully.
paid version just gets more and more complicated, all i want is to know how many days per week and miles per week i ran with my times trending up or down, but all the easy to use screens have been removed and replaced by designer mush. it’s like KPIs have gone wrong and they think the rage navigating is engagement. app was fine in like 2017. and then when they make me swipe through some product tour of overcomplicated BS or hunt for the concealed “close ad” button on some stupid year end recap when all i want to do is start my run. i think they have too many PMs competing on vanity metrics to justify their existence
Readit News
And I wonder if the article is correct in that they are pivoting away from segments. That seems strange since segments seem like the most compelling feature.
Route building and personal heatmaps are some of my favorite features, but there's so much that could be added (like a simple search). To some degree their customers are spread out over so many market segments beyond the original biking and running that they started with. Each segment has their own needs and wants. Then there's all the hardware (watches, power meters, trainers, etc..) that needs to be integrated and supported.
I posted it on HN but got flagged[1]. I still find it eye opening, a reminder that the code we write are products impacting the finite life of other people. Maybe we should have more modular, more open software as a standard so that other people can work on and deliver experiences that the original creators fail to deliver.
[0] https://web.archive.org/web/20200601152323/https://support.s...
[1] https://news.ycombinator.com/item?id=14935487
I used it snowboarding and there's some way to track runs and push out the whole day of skiing as one post, but I couldn't figure it out, so I just polluted my own feed with these single run posts. Then for some of them I forget to start recording halfway down the run because once again thanks to the broken auto pause feature, so now I need to remember to start it up at the top of the hill in addition to already having to put on a snowboard binding. Then thanks to the auto pause and me forgetting about pausing the run, it records me shuffling about in the lift line and then riding 25mph on the nose on the gondola.
Most apps/startups begin as an exploration of product/market fit. They'll try a lot of different things and use analytics (even simple server-side stats) to determine what people actually use in the app. Very frequently, you discover that the things you thought your users would want are actually only used by 0.05% of your customers. Eventually you have to start shedding rarely used features and limiting free plans, even if it makes the non-paying users angry.
Truth is, it doesn't really matter if you're losing someone who spend 5 years on the free plan but refused to sign up for the paid plan. They're not converting to paid unless they're forced to, and you're not gaining any money by letting them stay on the free plan for another 5 years.
> but where are all the new and cool features that should have come out over the last few years?
Cutting rarely-used features like this one could be a sign that they're trying to free up engineering resources to ship new features.
This is true, but it ignores the importance of network effects in products based on social media/interaction. I'd guess that a huge part of the attraction and user retention ability of Strava is the social aspect, which they'd be crippling if free users migrated somewhere else. As a secondary issue, Strava also benefits from the data generated by free users, though I have no idea what the value of that data might be.
They paired back things that weren't working, moved to focus on revenue (as they weren't going to be able to grow into the social network for all activities like they tried).
I think it's the right strategy, but this move makes no sense. It's very easy for them to manage duplicate data, and I'm surprised to see them suggest users set-up their garmin device to connect directly to Apple. This essentially makes it easier for strava to be cut out of the loop in the future if the social aspect isn't working for you.
My normal morning jog goes under a large bridge. This often confuses my phone's GPS, leading it to return a spurious point or two a kilometer away. Strava is completely incapable of detecting these spikes (accelerating to 200 km/h when running is normal right?) and offers no way to edit them out afterwards, meaning all my speed and distance records, progression charts etc are ruined by junk data.
That's not what DCR wrote, that part was clearly about user perception: Strava has grown into something bigger than just that app that declares you KOM when you go really fast.
Deleted Comment
And it apparently isn't valuable enough to charge more for ($5/month)
What are you referring to? That Strava has to store&compute more and more data from a user the longer they are using the platform?
The Strava/Garmin relationship is an interesting one: Strava has the social network, Garmin has the best devices [for serious Athletes, not casual users]. Garmin Connect is pretty cool in it's own right, but the Social features never really took off, which is where Strava plays and has a de-fact monopoly. Strava can't survive without Garmin, and Garmin benefits from Strava's content.
Garmin has pulled some 'power moves' in the past though with people it doesn't like... A competitor: Wahoo, who made cycling GPS computers, was cut off from inserting data into Garmin Connect and it left a lot of users out in the cold. Most serious cyclists will use Garmin devices, not an Apple watch, to track their rides as it seamlessly integrates with ANT+ sensors: power, cadence, wheel speed, heart rate, chainring and cog positions sensors.
Interesting to see Strava cut off Apple... I'm guessing it has to do something with preventing them from developing an alternative to the Strava social network.
I guess their huge data breach that they didn't notify and spent weeks fixing doesn't phase you, but it punted me out of their system.
> Garmin is the only one I actually trust currently, as they've shown 0 willingness to monetize your health information.
What about the concern that the company is bought out in the future? Or that it may be sharing data already, with government agencies, etc? Or companies that it works with?
I honestly can't even imagine the criteria whereby a corporation could be 'trusted' with personal information!
It boggles my mind, that people think about this - and plainly you do - but come to the judgement that its ok for corporations to have this personal data!
> What about the concern that the company is bought out in the future? Or that it may be sharing data already, with government agencies, etc? Or companies that it works with? > It boggles my mind, that people ... come to the judgement that its ok for corporations to have this personal data!
I'm not OP. I chose Garmin precisely because I don't think it's ok for corporations to have this personal data.
I chose Garmin because I can use it without needing to share any* of my data with a third party, including Garmin itself. I save the workout files to my computer via the same USB cable I use to charge the watch. There are various non-cloud apps I can use to view and analyze these workouts, if I actually cared to do so beyond the "fastest 5k" etc that the watch tracks automatically.
For this, it does not matter to me whether Garmin is bought out in the future because my watch works just fine as-is, and cannot update its OS without my explicit permission. I'm unsure what data Garmin would be capable of sharing because I have given it none.
* of course, speed/time/location data is obtained by pinging GPS/GLONASS satellites, but the watch can and does record my workouts quite accurately in non-GPS pedometer mode.
Apple had a chance to make on-premise data storage a thing. They had AirPort, which combined with a Mac Mini could provide full iCloud functionality from the home. iCloud itself could have simply been an encrypted offsite backup, like Tarsnap with no options.
Instead Apple built a system that still puts clear user data within their reach.
To me, it comes down to me being fine with this data being considered public. If others don't want to share it, that's totally fine and I understand their perspective. For me though? I couldn't care less.
Congrats, you sold some data that says I'm fat, out of shape, and don't like riding my bike on busy roads.
I’m a happy Garmin user but I know a lot of distance runners that have switched to Coros. They love the battery life w/o sacrificing GPS accuracy.
Strava -> Apple Health, still works
Garmin/Fitbod/3rd party app -> Strava -> Apple Health, no longer syncs.
IMHO this is better for me because now I don’t have duplicates in Apple Health and can sync to both services. But to each they’re own I guess.
BTW and a bit OT, I find it very impressive that Strava can retroactively create a leaderboard going back years for a newly created segment, meaning that they must evaluate potentially millions of nearby activities for overlaps, often in just a few minutes. That's a hell of a query. Anybody know of more information on how they do it?
Even if it is a couple orders of magnitude larger than I think geographic partitioning can keep the volume small enough to easily fit in RAM.
Deleted Comment
I was actually looking into this myself, and a few parts of strava are not that hard. For me the overview of my rides (including avg speed and distance is important), I quickly implemented a webpage with javascript and leaflet (map viewer), and I could present that data pretty quickly using geojson. I had an offline program which converted gpx traces to geojson, but I'm pretty sure I'm able to read gpx in javascript directly. The harder part is generating the gpx trace (which I used komoot app for), as this involves matching your trace (gps positions) on a map instead of using the data as is.
And after all this, your user base consists of the people who didn't want to pay for strava.
But a self hosted tool still can do a lot for own analysis. For tracking the own progress, the own training plan.
Progress is slow as a solo dev who contributes to it in their spare time, but I’m happy enough with it that I could archive my Strava account and switch to it fully.
I guess you might be able to do interesting dashboards with grafana.