City Guide App Store Preview Video


tl;dr:

Three weeks after releasing an App Store Preview Video in the Apple App Store for City Guide, we saw a 4.76% increase in page view to install rate overall, and a 40.39% lift in page views to installs, for traffic from Paid Twitter, among users who we know definitely saw the video.

Experiment Overview

  • App= City Guide

  • Tested in Apple App Store

  • Name = City Guide App Store Preview Video iOS Experiment (not recorded anywhere except in this doc.)

  • Start Date = 4/23/2018

  • Measurement: Look at app store page views and installs before release, and after release.

  • US only


First, we looked at the page view to install conversion rate of City Guide three weeks before the video was released in the app store and three weeks after. Limitations: This approach does not allow us to measure only users with devices where the video would appear (iPhone 6 and higher.)

app store preview video data 1.png

But because the app store preview video only shows on iPhone 6 and higher, we needed to take a more granular look at the impact the video had on the people who could view it.

We dug deeper and looked at installs and app store page views by screen size for two weeks before, and after the app store video launched. As noted above, we have limitations in seeing page views from the screen size iPhone 6 and higher from all cumulative sources. However, we can see that data coming from Paid Twitter, our biggest paid acquisition channel. Here is the impact on traffic coming from Paid Twitter:

App store preview video data 2.png

Results:

Three weeks after the video was released, we saw a 4.76% lift in page view to install rate (99% statistical significance). This could result in 1,413 additional monthly installs, or 47.12 additional daily installs (US only). When drilling deeper to a subset who we know definitely saw the video (Paid Twitter, iPhone 6 and higher), we saw a 40.39% lift in installs, which could result in 186.15 more installs monthly (99% statistical significance).

Next Steps:

  1. Consider creating an app store preview video for Swarm iOS or Swarm Android.

  2. If iOS measure results with the same approaches, if Android, run a clean A/B test through the Google Play Console.

  3. Compare to these City Guide iOS results.

(And yes, I did pick out the music in the video! Praise ye, Audio Jungle!)

Erika Finnimore