Segmentation in Incrementality & App Retargeting | AppsFlyer
4 Min. Read

Incrementality & App Retargeting: The Importance of Segmentation

Avatar Shani Rosenfelder Aug 30, 2018

Since the dawn of time (or at least since advertising began to take shape), marketers and their bosses have been asking themselves “How do I know that the marketing investment has actually generated real value?”

Their concern makes sense. After all, the fact that a campaign performed well does not necessarily mean it justified marketing spend. The million dollar question is whether the paid-for users who generated revenue would have done so organically, which basically means for free.

To help answer that question and alleviate those concerns and justify spend, incrementality tests surfaced.

What is incrementality?

Incrementality is the measure of the lift that can be attributed to a marketing campaign. In one common method, marketers compare a baseline of expected/organic activity and then isolate the benefit of marketing on that activity. The following chart explains this concept:

One specific method in an entire discipline of testing for incremental effects of ads is through A/B testing. In this test, the impact of an ad served to an exposed group is compared against a control group which is not served with that ad or is served with a ghost ad/PSA (public service announcement ad).

Enter retargeting

The notion of incremental revenue is particularly important in retargeting campaigns. In this scenario, users have already installed and used an app, so the likelihood of them engaging with it again on their own volition (organically) are much higher than the chances of a user organically discovering an app for the first time. Hence, the question that lingers in the back of marketers’ minds: to retarget [through paid campaigns] or not to retarget, and to what extent.

To help our clients understand the incremental value of their retargeting campaigns, we have recently conducted several A/B tests. We’ve found that, in most cases we analyzed, retargeting did produce incremental value.

Having said that, it is important to stress that there were instances where retargeting did not only fail to deliver incremental revenue, it even detracted from what would have been generated organically. This was most likely the result of a poor user experience driven by bombarding users with ads, which led to resentment and ultimately reduced spending!

In retargeting, segmentation is the name of the game

By analyzing the campaign of one shopping app, a clear difference quickly emerged: retargeting the segment of users who have yet to complete a purchase generated significant incremental revenue, while the segment of those who did purchase pre-exposure had a substantial negative impact.

We can see that for this app, re-engaging users who have already made a purchase is not recommended since it may deter those shoppers from buying again. Is your retargeting campaign doing the same? Make sure you check it.

On the other hand, the retargeting campaign did wonders for the segment of users who did not purchase before being exposed to the campaign, with substantial lift beyond two days from the install. A month in, retargeting proved extremely beneficial, generating significant incremental value.

What this means is that this app would be better off refraining from running retargeting campaigns during the first half of the week after an install, while allocating the lion’s share of its budget to retarget users only during the fourth week after the install.

Another example from another shopping app yielded a clear result: retargeting activity for the set segments and goals was highly incremental from day 0, particularly as far as converting users goes. This segment did not purchase before being exposed and thanks to the campaign its users converted at far greater numbers than the control group who was not exposed to ads — increasingly so over time reaching a lift above 5x during the 4th week since the install.

The campaign also had a positive effect on the number of lapsed users, effectively reducing the number of users who ceased purchasing activity. Similar to the lift of converted users, this trend only increased over time.  

Make sure to properly monetize your converting users

This gaming app managed to generate tremendous lift in the number of converted users by running a retargeting campaign. However the average revenue from these users did not perform as well, and even detracted from the organic baseline in certain cases. A more significant revenue uplift was recorded in the 4th week after the install.

In this case, we can see that the retargeting campaign delivered mixed results and the marketer is seeking ways to generate more revenue from these converted users.

Don’t forget to test networks!

In addition to testing different segments to pinpoint those with the highest incremental lift, it is equally important for marketers to test multiple retargeting networks. The following example from a 3rd shopping app whose performance was analyzed shows that different networks delivered varying incremental lift for the same segment — loyal users who made a purchase before and after being exposed to a retargeting campaign.

The bottom line

Understanding incrementality allows advertisers to optimize their ad budgets by allocating retargeting ad spend in the most impactful way possible — using the right segments, at the right time and across multiple proven networks. Indeed, we’ve seen that in most cases we had analyzed, retargeting certainly proved its incremental value.

Part 2: Dormant Users

The above analysis was based on UA users and the time between the install and the exposure to retargeting. Stay tuned for Part 2 in the series where we will focus on dormant users and analyze the time between the last engagement/purchase and the retargeting exposure.