Uber has seen a sharp drop in retention rates for new drivers in the U.S., according to analysis of the Uber driver app provided to TechCrunch by app analytics firm Apptopia.

In an analysis of app downloads and usage, Apptopia estimates that 30-day user retention for the Uber driver app in the U.S. has dropped 47 percent from January through May.

This measure looks at the proportion of users opening the app each day after the initial day of download continuing until the 30th day the idea being to measure engagement meaningfully versus looking at app deletions (as lots of people just stop using an app versus actively deleting it).

Apptopias analysis also indicates a 20 percent bump in downloads of the driver app over the same period.

So if the data crunching is correct it appears that while Uber is successfully managing to drive initial interest from new drivers, its having serious trouble sustaining this interest.

From April, retention rates appear to fall especially dramatically.

Apptopia does not get any usage data direct from Uber but pulls data from a network of250,000 apps to which it has developer account access. It then uses an algorithmic framework to generate estimates for individual apps, including by using public signals such as App Store reviews and says its methodsresult in strong trend data for major apps.

For the two percentages its pulling here, it says its averaging its data from the Google Play Store and iOS App Store together for, as it puts it, a more holistic view on interactions with the Uber driver app.

We asked Uber if it had any comment on the data, but at the time of writing it had not responded.

The companycurrently has no CEO in its own driving seat, after co-founder Travis Kalanick resigned following investor pressure applied in the wake of a report into its internal culture triggered after a female former employee blogged about experiencingsexual harassment and sexism during her year at Uber.

Its unclear whether new Uber drivers are sensitive to Ubers internal turmoil. Perhaps more likely is general dissatisfaction with lower rates of pay from Uber pool rides and Ubers lack of an in-app tipping feature (versus Lyft having in-app tipping) tellingly,this week Uber finally said it will start allowing riders to tip drivers via the app.

Asked for his view on Apptopias data, Harry Campbell, founder of The Rideshare Guytold us: Uber drivers for the most part have been very happy this week because of the tipping option and TKs [Travis Kalanicks] departure. Weve known for a while that Uber has problems with high turnover and low satisfaction rates amongst drivers and a lot of drivers felt that TK was the root of many of their problems.

Apptopia also looked at ridermonthly usage for us, and on this said perhaps surprisingly that Uber hasnt taken a hit due to #deleteuber i.e. the social media campaign that sprung up aiming to convince Uber users to ditch the app as a result of the various ethical scandals being attached to it.

It drove uninstalls but new downloads remained consistent so usage didnt dip much, said Apptopia of the #deleteuber movement. Since February, monthly usage in the US (for riders) has actually increased around 60%.

Engagement [aka the frequency of app opening per individual on average within a month] is down in the US, but really not by much.

Read more: https://techcrunch.com/2017/06/23/uber-has-seen-a-sharp-drop-in-new-driver-retention-this-year-apptopia/