When Do People Click or Dismiss Your Push Notifications?
At OpenBack, we are dedicated to helping mobile marketers determine the perfect moment of delivery for their push notifications. This moment – the moment when people have no other distractions and are most likely to click your notifications – differs on a user-to-user basis. (This is why micro-segment and micro-moment targeting is a key strategy when it comes to crafting your push campaign strategy.) There are many different facets to “the perfect moment,” however. The subtle mobile marketer must take into account not only the time of day, day of the week, but also all of the other contextual factors per users… They must add to the equation the other data they can use from that user and their device as well.
Analyzing billions of anonymized data points from push notification campaigns, the team at OpenBack has been able to analyze crucial insights on user behavior. Below, we have published our findings on what time of day and day or week people are click and dismissing notifications, for cohorts of 4 different regional clusters. We also include data sets on how device battery life, and whether a user’s phone is on mute or high volume, etc. affects interaction rates. From all of these insights, it’s clear that finding the perfect moment is more than just a matter of time of day and timing. Rather, it’s made up of many different contextual signals – including time of day, users’ geolocation, device ringtone volume, battery life – showing massive distribution of when people are likely to click or dismiss notifications
Download our NEW Metrics & Measurement Playbook to learn how to optimize your push campaign!
When Are People Clicking Your Push Notifications: Time of Day & Day of Week
We aggregated data for 4 different cohorts of users based on geolocation: US and Canada, UK and Ireland, Russia, and the Middle East. We then tracked notification interaction rates (by which we mean both clicks and dismisses) across every hour of the day. In the graphs above, we have charted average CTR per hour for the regions of US/Canada and UK/Ireland, as well as the break-down of CTR per hour for each day of the week.
Results were fascinating, and point to considerable behavioral differences between app users on both sides of the Atlantic. In the US/Canada cohort, interaction rate never hit ultimate zero, and even in the hours between 1am and 5am, interactions hovered slightly above 1.0%. This suggests that North Americans can’t quite manage to switch off entirely, not even when they’re sleeping. Interaction rates taper upward slightly around 6am – indicating this is when people are waking up, getting ready for work, commuting. At 10am, interaction rates plateau at around 5.0%, until they spike to 14% around 8pm.
There are slight variations according to days of the week: for example, on Saturday the spike in interactions (11%) occurs around 9:30am and then slopes downward, indicating that people are waking up and checking their phones, then putting them down for the rest of the day because they have more enjoyable things to do.
This cohort differed in a few key ways. For one thing, contrasting with the more even topography of the US/Canada cohort, interaction rates were more extreme. Interaction hit zero around the hours of 2am to 4am, then gradually increased to above 4% as people woke up and went to work. Then, in a complete divergence from the previous cohort, UK/Ireland interactions spiked to over 18% at 10am. This might speak to differences in work culture in the regions: in the UK/Ireland, people likely settle into work at 9am, get some housekeeping done, and then take a coffee break around 10am to clear their heads and check their phone. Whereas US/Canada might have more of a “hunker-down-until-the-end-of-the-day” ethos.
Interestingly, the graph on Thursday shows a second spike in interactions, and then a third smaller one at 6pm and 9-10pm respectively. This may be owing to the fact that in the UK and Ireland, Thursday is a popular night for going out.
Russia breaks from the other three cohorts in that there is no clear spike of interactions at any hour of the day. Interaction rates rise from 0.5-1% around 4am, to hit a high point of 7% around 9am. There is a slight downward trend for the rest of the day, with a sharp slope from 5% back to 0.5% between the hours of 8pm and 12am.
The interaction rate patterns across different days of the week are even more interesting, in that they are the most erratic of any of the cohorts. They all follow somewhat the trend of interacting with notifications more in the hours between 5am and 8am. However, there are so many peaks and dips for every day of the week that it’s difficult to draw any conclusions on when people are clicking on notifications, except that phone signal and/or moments people can check their phone are likely not very consistent in Russia.
Their highest rate of interactions (8am on Saturday) also barely clears 10%. Compare this to US/Canada’s high of 37% on Sunday, UK/Ireland’s high of 36% on Tuesday, and the Middle East’s high of 19% on Tuesday, and it would seem Russian users just don’t have a strong habit of engaging with notifications.
While Russia’s data was the most erratic, data on when people from the Middle East are clicking and interacting with notifications is the most uniform across all 7 days of the week, with a slight variation on Saturday evenings as interactions trend up instead of down.
Interactions are at a flat zero from midnight to 5.30am. After this, interactions spike upward for the high point of the day, an average of 10.5%. After that, interactions drop sharply and slope downard slightly, with a second spike around 6:30-7pm.
When Are People Clicking Your Notifications: Device Status
In addition to time of day and regional location, people’s likelihood of interacting with notifications can depend on the device itself. We have aggregated data on how battery life and device volume level affect interaction rates, as seen below.
Here, the data perfectly illustrates what common sense would suggest: most interactions (16%) occur when the device is at peak battery levels (96-100%). There is a steep drop to around 7% for the next percentile of 91-95%, and then a gradual downward slope to 0% interactions for 0% battery life. This graph is the perfect demonstration of why using OpenBack’s battery life signal can boost user interaction rate: simply waiting until they have charged their device to above 96% can more than double your average click-through rates.
Like the previous chart, the data on Ringer Volume Level shows the moment when people are clicking and dismissing notifications is highest at the 96-100% percentile, with clicks at nearly 45%. This drops drastically to around 5-6% for the 91-95% percentile. There is a predictable downward trend that for the most part coincides with how low users’ ringer volume gets. Then, at 0% ringer level interaction rates shoot back up to 23% clicks and 30.5% dismisses. This may be because some users set their devices to “vibrate” when they’re on silent, meaning notifications will catch their attention anyway.
If you are interested in learning more about OpenBack’s unique data signals, as well as our data insights on user interactions resulting from the use of these signals, get in touch with one of our experts.