As a CRO consultant this is interesting, but I'm skeptical. (It's not particularly your fault. Nearly every marketing tool promises nirvana, but very few of them deliver.)
It sounded great until I realized that:
1) I would still have to think of variations and create them myself.
2) Even after that, this would only be useful if things like location and time of day really did have a significant effect on conversion rates.
3) And even after that, this would only be _worth it_ if those effects were not obvious to me and could only be discovered with ML. For instance, it doesn't take ML to hypothesis that a visitor who came from an ad link (&utm_campaign=dogs) is more interested in seeing a page about Dogs. And that hypothesis can be tested for free with Optimizely.
In other words, ML is cool and all but I don't see what value this adds to conversion optimization.
Maybe my assumptions are wrong and I'm missing something, in which case I hope this is useful feedback.
1) Yes, the problem with creating variations is still here and is worth mentioning. Today you need to know your product well and know your visitors well to create a valuable hypothesis. And this is what we're working on right now (disclaimer - I'm one of the co-founders of Landy).
2) The idea here is that with ML you should not analyze every dimension separately. ML is taking into the account all available characteristics and making decisions based on all of them together (like if the guy on OS X, who came from NY from the Facebook campaign in the evening - prefer to watch product video instead of watching screenshots - no problem, we'll show him video).
3) The real power of ML comes out when you could not obviously split your traffic based on the ad link (like utm_campaign=dogs). Direct and search traffic on your homepage are great examples in this case. Also, manual targeting requires a bunch of analytic folks, who will continuously analyze your traffic, setup and adjust optimization campaigns. Even in this case - it's still difficult to adapt to dynamic changes in traffic (like a new type of visitors, season changes, etc). So ML could not only improve results but also decrease the amount of human resources which is currently required for solving such complex problems.
So what I'm trying to say is that your assumptions definitely make sense in some cases. But we believe that there are still plenty of cases when ML could drastically increase your results and save your time.
I had a similar idea but couldn't convince people on my team to try it out. Glad to see someone is trying this. Adding a SaaS model is a good idea.
Have you tried switching out parts of a page? I tried switching react components in an admin interface. It took into account the user id, so in some cases, the user would get a 'customized' UI.
I had trouble tuning the data to make the output actually useful, so the project has been shelved for now. Hope this works out for you guys.
Actually, it is. We're currently running simple personalization campaign with two versions of landing pages which has different design, messaging, etc.
This makes sense. Ads have been personalized for a long time so why not pages? Some work is needed to create these but for people trying to squeeze every conversion they can out of visitors is may be worth the trouble. It also might help prevent fast bounces.
Readit News
It sounded great until I realized that:
1) I would still have to think of variations and create them myself.
2) Even after that, this would only be useful if things like location and time of day really did have a significant effect on conversion rates.
3) And even after that, this would only be _worth it_ if those effects were not obvious to me and could only be discovered with ML. For instance, it doesn't take ML to hypothesis that a visitor who came from an ad link (&utm_campaign=dogs) is more interested in seeing a page about Dogs. And that hypothesis can be tested for free with Optimizely.
In other words, ML is cool and all but I don't see what value this adds to conversion optimization.
Maybe my assumptions are wrong and I'm missing something, in which case I hope this is useful feedback.
2) The idea here is that with ML you should not analyze every dimension separately. ML is taking into the account all available characteristics and making decisions based on all of them together (like if the guy on OS X, who came from NY from the Facebook campaign in the evening - prefer to watch product video instead of watching screenshots - no problem, we'll show him video).
3) The real power of ML comes out when you could not obviously split your traffic based on the ad link (like utm_campaign=dogs). Direct and search traffic on your homepage are great examples in this case. Also, manual targeting requires a bunch of analytic folks, who will continuously analyze your traffic, setup and adjust optimization campaigns. Even in this case - it's still difficult to adapt to dynamic changes in traffic (like a new type of visitors, season changes, etc). So ML could not only improve results but also decrease the amount of human resources which is currently required for solving such complex problems.
So what I'm trying to say is that your assumptions definitely make sense in some cases. But we believe that there are still plenty of cases when ML could drastically increase your results and save your time.
Have you tried switching out parts of a page? I tried switching react components in an admin interface. It took into account the user id, so in some cases, the user would get a 'customized' UI.
I had trouble tuning the data to make the output actually useful, so the project has been shelved for now. Hope this works out for you guys.