Benchmarks deep-dive II: Paid versus organic traffic
Are users who come to apps after clicking on an ad more difficult to retain than users that were actively acquired? Or do these users stay in the app longer, perhaps because the ads prepared them for the in-app experience, providing a smoother user journey? How often do users of these groups use their apps each day? And are Android users different from iOS users, like we found for some of the Games verticals? Let’s take a look!
Organic Android users show the best retention
The first engagement statistic we will look at – just like we did in the extensive Mobile Benchmarks Report for Q3 2015 – is that of pure retention: the rate of users that come back on each day-after-install.
The first point of note is the close proximity of the groups we’re observing: there are only minor differences between Android and iOS, or paid and organic users. We can see that on day 1 after-install less than one in three users return to the app on both Android and iOS. Organically acquired Android users are moderately more likely to use the app again on this first post-install day than acquired users are: 31% of organic users return on this day, in contrast to only 27% of acquired users. The retention rate for acquired users is virtually the same for iOS, and while organic users retain slightly better at 28%, the most noticeable difference is between organic Android users and the rest here.
The development over the next weeks shows that in the first week, many users continue to reduce their app usage. By day 5 after-install, the day 1 retention rates are cut in half, and by day 12 only one third of the users who returned to the app on day 1 after-install are still active.
Notice that there is little difference between this day and Day 14. We observed the same curve trends on day 7, and see them again on day 21. The trends are curious and could simply be cyclical: such that some apps might see higher install numbers on weekends, and their users would then shift the picture slightly, showing a higher retention once the next weekend rolls around. But we think that these upticks around day 7 and 14 signify the success of engagement campaigns aimed at dormant users after a few days of inactivity.
Interestingly, after the first week iOS users pick up the pace – relatively speaking. The remaining 11% of organic iOS users active in the app on day 8 after-install retain better over the next weeks than the 12.5% organic Android users from the same day. iOS users are also more strongly affected by the day 14 and day 21 bump effects discussed above.
The same trend with iOS users being slightly more loyal to their apps once they have accepted them can be seen in paid traffic: Over the first five days, acquired users on Android and iOS drop off at fairly equal rates, but after this the Android users fall behind a bit.
Time spent per retained user
Let’s take a look at how much time users spend in the app, depending on whether they were acquired or organically came into the app.
Curiously, iOS users and Android users show opposing patterns for time spent in-app depending on whether or not they were acquired. Just as with the last benchmarks deep-dive on the Games verticals, we see that the trends stabilize after the first few days. During these days iOS users behave very similarly, but eventually we see that the acquired users actually use the app longer on each day than organic users do. On Android, however, the acquired users spend markedly less time in the apps than the organic users right from the start.
Note that this data only compares the time spent in the app by users who actually used the app on the given day. This shows that the overall time spent in the app goes down, users become less engaged with games apps and at the same time learn to use their utilities apps better, requiring less time spent in the app. Take a look at the benchmarks report by vertical to see which genres might contribute to either trend.
Ad-acquired iOS users spend more time in apps than organic users. - Tweet this!
Paid Android users clock in at more than 20 minutes in the app on their install day, with the average usage time eventually dropping below 15 minutes by the second week. In contrast, the organically acquired users spend less than 12 minutes on average once the patterns stabilize. Note that these effects might well be due to a stronger games focus of the Android audience, possibly along with some app groups like podcasts that are often used more on iOS.
Sessions per user
Finally, take a look at the following graphs showing the session count per active user:
The overall pattern we find here is very similar to the one for time spent above: after an initial drop, the curves find a certain horizontal center around which they revolve and from which they don’t often diverge. Notice that the overall picture is paramount here, as the differences between the curves are fairly small. The average session counts for engaged users remains between 1.8 and 2.2 sessions per day for the entire period of 30 days and for all four groups.
We nevertheless see that the difference between acquired and organic iOS users is very small, with the organic users curve beginning above and ending below the acquired users curve. On Android, we again find that organic users outshine paid users, who trigger about 10% fewer sessions per day.
Back to the big picture
So, what are our key takeaways from all this? Overall, we’ve found that there is a marked difference between Android’s paid and organic user behavior, and that the iOS user groups are more homogeneous across acquisition types.
Non-organic Android users show lower engagement stats than the Android users acquired organically in all three observed metrics: the ad-acquired users continually have lower retention rates, spend less time in the apps, and also open the app less often.
In contrast, there are only small differences between the different iOS users. Organic users retain better in the long-run, but only slightly. However, their daily time spent metric is lower for those later days, and the trend on the session counts also favors the ad-acquired iOS users.
How does this make sense? Working from our earlier hypothesis of paid traffic being generated mostly in Games, it could be that the retention drop-off occurs sooner for retained games users than for users in other verticals, but that these users are also the ones with the higher values on the non-retention engagement KPIs, such as time spent and number of sessions.