Retargeting vs. new user acquisition: Which method wins?

This article originally appeared on Adjust’s blog, here.

Retargeting works, at least that’s what we set out to prove when we started with this study. Our hypothesis was simple: that retargeted users outperform new users when it comes to retention, events and (most importantly) revenue events. Below is a look at what we found.

We won’t waste time explaining ‘what is retargeting’ in detail — after all, you probably well know what it is. That said, if you’d like to improve your knowledge about retargeting, you can find a full definition in our glossary, and even more in any one of these articles.

Key takeaways

In our data, we’ve seen that there’s a statistically significant difference between retargeting and acquisition in their average performance aggregated at both per week, and per day.

While new users have more sessions, engagement of retargeted users is noticeably improved in terms of the number of events they trigger, and they perform slightly better in terms of revenue events and retention rate. Older retargeted users have a higher engagement rate because they already had an experience of using an app. We explore this rationale in more detail further below.

Three-point summary

  • Retargeting campaign users have an overall 152% higher engagement rate (number of events per user) than new user acquisition campaigns over 30 days. On day 1 of install, it’s nearly 200%, before dropping off significantly by week 4.
  • Retargeting campaign users also make 37% more revenue events in the first 30 days than new user acquisition campaigns.
  • On the first day of a campaign, retargeted users retain by 5% more than new users, and typically remains at a 5% difference by day 7.

Methodology

We used four cohort KPIs to compare between two samples (those being new users and reattributed users.) All data were collected for the time period between January 1st and July 1st, 2017. The cohorts are:

  • Number of sessions per user
  • Number of events per user
  • Number of revenue events per user
  • Retention rates

Of our sample, we tracked 140 apps that track with Adjust. Between them, there were over 500 retargeting campaigns, versus nearly 9000 acquisition campaigns. Reattributions and retargeting campaigns on either side tracked into the millions, though the sample of installs (as with campaigns) for initial acquisition campaigns was 10x that of reattributions.

Events

Events are the actions performed by users in-app. Whether it be completing a level, making a purchase, or even clicking a link, any action that takes place within an app is known as an ‘event.’ Let’s look at how they differ depending on user type.


Users from retargeting campaign have a noticeably higher engagement (number of events per user) than new user acquisition campaigns. This is the start of a trend you’ll see, but with events in particular, retargeted users consistently outperform their newbie counterparts.

On day of install, it’s near-double, with 86 aggregated events per user in a single week vs. 44 for new ones. This is quite interesting to note, as users would have already been through before, so it might suggest that users who are more comfortable with the experience may be more keen to continue with a deeper experience on the first day, avoiding onboarding fatigue.

By week 4, there’s been a significant drop-off, though retargeted users still complete 10 more events in a single week than new users do. If you’re tracking with Adjust, you may want to take a look yourself and see if our numbers are consistent with yours.

Revenue Events

As with events, retargeted users also make more revenue events in the first 30 days, and also further, up until about week eight, whereafter both types of users behave relatively the same.


The numbers themselves are much narrower than with events — here we see a nine percent difference on day one of install. However, by day seven, it’s a 49 percent difference, where retargeted users are still likely to trigger a revenue event. That’s a large outperformance.

So, from what we can tell from here (and over the 12 week period graph seen below) is that retargeted users trigger more revenue events, therefore generating more revenue.


LiftOff’s latest acquisition data shows that “the average cost to acquire an app user who makes an in-app purchase — like game content — is $76.40,” so prioritizing retargeting campaigns that successfully engage and drive revenue can be a method to decrease this cost of acquisition.

If retargeted users make more purchases (at a lower cost of acquisition), not only is spend saved, but lifetime value (LTV) is increased for retargeted users, making them a much more interesting prospect if you’re not already engaging with your lapsed user base.

Of course, for users to become retargeted, they must start out as new users. What this means is both audiences are important, and it’s vital that you don’t neglect new users who’ve since abandoned your experience — as they’ll likely present great targets to produce more revenue, and more engagement.

Retention

In general, retargeting campaigns’ cohorts have a slightly higher level of retention than that of new users. Taking a look at the chart below, which displays retention for the first 30 days of install, on day one retargeted users retain by five percent more than new users.


This gap fluctuates wider within the first week, and remains at a five percent difference by day seven. Beyond that, the two narrow in difference over time, with reattributed users always remaining in higher numbers for longer. Retargeting campaigns’ cohorts have a slightly higher level of retention rate in the first 6–8 weeks too, but then they become the same.

Retargeting works!

If there’s one main takeaway from our study, it’s clear that retargeting keeps users around and engaged, and also (most importantly) drives increased revenue while lowering cost of acquisition, especially within the first few weeks of an install.