November 20, 2012
Push notifications are a popular tool for this purpose. With the vast majority of app revenue growth coming from in-app purchases rather than paid downloads, it’s more important than ever to keep users engaged with your apps. Updating users about new content or activity and driving them back to the app. At the same time the technology is controversial since unwelcome or excessive push notifications annoy users, who will either disable them or worse yet uninstall the app responsible. So is it worth the risk? Do push notifications work?
It’s clear that push notifications are better suited to some app categories than others; new emails, instant messages or updates to social feeds are the sorts of things lots of people would like to be notified about. Letting a player know when it’s their turn in a turn-based game is another obvious use case, while other valid gaming uses are less clear. There is absolutely no doubt that push is driving significant engagement for some developers. However, real data on the effectiveness of push notifications across a wide array of apps is hard to find. Push provider Xtify has revealed some top-level engagement metrics :
“30-60 percent open rates and 4-10 percent interaction rates (with spikes as high as 40 percent).”
Purely for comparison, this is significantly higher than the level of engagement that could be expected when sending emails (average open rate 20 percent, click-through rate 5.4 percent), although not as high as SMS (studies vary but open rates are typically >95% and click-through rates for URLs where included is often in the high teens). What push providers are not generally publishing at the moment is their equivalent of the unsubscribe rate (disabling push or uninstalling the app). It should be noted that app uninstall rates are high anyway, push or no push. This doesn’t seem to be directly related to app quality, so it wouldn’t be a fair comparison.
That’s averages, what if you use push well? Popular push provider Urban Airship provide an extensive set of resources to help developers create “Good Push”. Unfortunately the initial study they published on the results of “Good Push” is useless. It showed significantly higher engagement and retention for their top 10 users of “Good Push”. However, the top 10 were determined by having the best ratios of opt-in to opt-out (from push notifications) user engagement and retention. There are two problems here:
Fortunately Urban Airship have just published a follow up study with a much greater sample of their users. Sadly it still suffers from point (1) above in choice of metric but now at least we have data about apps that have much lower engagement with push notifications. Even where push notifications are clearly not essential to the app (less than half of users opt-in), the retention after 4 months of users who opt-in is ~80% greater than of those who don’t (34% vs 19%). Additionally, across all levels of engagement, a significantly higher proportion of app opens come from those who have opted-in.
This is the most useful hard data available. We can conclude from this that opting-in to push notifications is a good indicator of both retention and engagement. Having a direct communications channel with your most engaged users is an extremely valuable thing, since these are the ones most likely to convert to paying customers.
For now the advice is clear, push notifications are a valuable communications channel to your most engaged users – use them where appropriate and experiment with them to find out what works.
To keep users engaged with push a concise summary of best practice is:
If you use push notifications in this way to serve people with relevant and timely information then you’re much more likely to build trust and deliver something users value enough to pay for than if you view them as an alternative marketing channel on mobile devices.