landing page, onboarding sequence, and the initial out-of-box product experience are critical
Back to my math argument, which is totally an aside but I'm feeling pedantic: some things have compounding benefits, which I think is the real ripple. But the landing page does not. What the landing page does have is that it's really cheap to improve.
strong understanding of your user lifecycle
Well yeah, we gotta get that too. Do we have a graph of our retention funnel?
In front of the wall means features that create value without much investment, such as browsing a feed, rating some photos, or clicking a link
Or browsing a bunch. An argument for Firefox Search to import history from other sources? (Note, Firefox Search will work great right away for already retained users)
where “behind the wall” means that the feature can only be experienced once the users buys into a product, and engages
An aside to these discussions, but a good argument for why we should make features FxA-optional when possible.
Similarly, a product’s onboarding experience can be weak if there isn’t a strong opinion on the right way to use (and setup) the product
BAM!
A “day 7 feature” will automatically be used less than an experience tied to onboarding, since the tragic curve above shows that fewer than 4% of visitors will end up seeing it.
This is where the math doesn't make sense.
But there are other arguments. A D7 feature might be harder than an onboarding feature (especially an onboarding feature like messaging).
But maybe a better argument: a D1 feature is also a D7 feature, and a D30 feature. The earlier you retain users the more impact your retention can have.
Too little impact is made when they do engage. Especially the case when important/key functions are displayed like optional actions outside of the onboarding process.
I might posit: one reason these features have little impact is because the potential user is there for a reason, and your feature isn't the reason. And why would it be? Was your feature the first thing the user sees, that then motivated them to continue down the funnel? No, they saw the product defined in a general way.
That's why I think segmentation is interesting: we've started the funnel out with something different than "I want a browser" (who even asks that?) We've started the funnel with: I want help to better use the web to do X. Now we know what kind of value the potential user is looking.
the features target engaged/retained users rather than non-users and new users
This is the important concept, maybe the most important in the article.
A few weeks ago, I read this tweet, and found myself nodding my head in vigorous agreement. For people who love to build product, when something’s not working, it’s tempting to simply build more product. It leads to the launch-fail-relaunch cycle that I alluded to in a previous essay, Mobile app startups are failing like it’s 1999. However, this rarely works, and when you look at the metrics, it’s obvious why. The metrics behind the Next Feature Fallacy Let’s go into the numbers. The reason why the Fallacy is true can be described by one simple diagram, which might be described as the most tragic curve in tech: Screenshot 2015-05-31 19.50.54
The point here isn't that nothing can affect the curve, but rather the nature of the curve.
I don't think his implied analysis quite makes sense – if you are doing something that is OK delivering 20 people at D30, and you can decrease attrition by 10% from D1 to D30, then you get 26 people at D30. If you decrease attrition from first visit to signup by 10% you also get 26 people at D30, even though you've "gained" 80 people that have signed up. Except for some reason the math would lead to 28 people at D30 (D30=Signup/10), for reasons I haven't figured out. I guess because you lose 4 in 5 people from Visit to Signup, but you only lose 3 in 4 from D1 to D30.
All that presumes that it's somehow just as easy to effect a 10% improvement one place as another, which is unlikely. What is likely? Not sure.
This is a global note
threaded conversations
Are they threaded, or just a series of comments? That is, do comments form a tree?