Expert Answer
Anonymous
First I would ask, what is the motivation behind wanting to measure user growth and retention? At Uber, our mission is
reimagine how the world moves. For Uber Rides, it is to move people from Point A to Point B. In both cases, scale is an important part of the Uber mission, and it also contributes to Uber's revenue streams. The more riders there are, the more money can be made from subscriptions, service fees, and commissions. Also the more riders there are, the more demand there is for drivers, and the more need there is for drivers to join the app in order to keep supply-demand balanced.
So it will be critical for Uber to monitor user growth and retention, in order to keep the market balanced, as well as to make financial forecasts, and to evaluate the impact of experiments that are designed to augment those metrics.
Metrics for user growth include app downloads, sign ups, setup (filling out account details, adding credit card, etc.) and booking a first ride, perhaps within the first 7 days of downloading the app.
Metrics for user retention include
* use recency: DAUS, WAUs, MAUs
* use intensity: > $100 in rides in past week
* use patterns: ratio of DAUs to MAUs
* churn: % of users who uninstalled the app within the past month
Those metrics can all be computed on a cohort-basis, e.g., with weekly cohorts.
The metrics can also be sliced and diced, e.g., by market, by platform (iOS / Android), for riders versus drivers, and for age and gender.
These metrics would do a good job of capturing acquisition and retention