Beyond the Click: Mobile Retargeting Reality Check

Kochava integrates with self-attributing networks (SANs) like Google and Facebook via a suite of APIs. Integration largely comes down to campaign and tracker setup. Make sure that the settings configured conform to what you want and are not the MMP’s defaults. Important settings include lookback windows and minimum time to reengagement.

While you’ll have somewhat less control over your campaigns in large platforms like Facebook and Google, you can still configure critical settings within your MMP. The principle is straightforward—apply the same rigor to these major platforms as you would to any specialized retargeting partner.

In essence, reengagement attribution configurations are built to match clicks and events accurately using controlled time windows and event selections. Optimization involves tuning each of these parameters to suit your campaign dynamics, thereby improving the tracking and conversion of dormant users. For Kochava-specific mechanics of campaign and tracker setup, please refer to this guide.

The key is ensuring that your configuration properly reflects strategic marketing goals rather than just defaulting to the most generous attribution settings. Most importantly, don’t change everything at once—methodically review and adjust one parameter at a time, measuring impact before moving to the next lever.

Within Kochava, our omnichannel attribution solution collects and correlates user interactions across web, mobile, and CTV platforms to build a unified view of the customer journey. It tracks key touchpoints using controlled lookback windows and sophisticated identity resolution, ensuring accurate pairing of ad impressions and clicks with in-app events. This detailed mapping reduces attribution noise and confirms that mobile reengagement efforts target genuinely dormant users. As a result, our system optimizes ad spend, refines audience segmentation, and ultimately enhances conversion rates for your mobile reengagement strategy.

This comprehensive visibility helps marketers determine where and when a user truly disengages, allowing for more precise targeting of genuinely lapsed users vs. those who may still be active via other channels.

We frequently see concerning levels of inventory overlap, with some users receiving impressions from multiple retargeting partners simultaneously—essentially causing advertisers to bid against themselves. Your MMP data can identify this problem, but solving it typically requires direct coordination with partners to divide targeting responsibilities more effectively.

Frequency caps and suppression lists can be useful but do have limited effectiveness, especially on iOS due to ATT.

As noted during our discussion, “Anybody who’s been in this game for long enough knows that frequency caps are very, very hard to manage, especially on iOS.” The ability to maintain an accurate suppression list has become increasingly difficult with the introduction of App Tracking Transparency (ATT), especially for targeting iOS devices. On Android, frequency caps are easier to manage.

It’s essential to verify that whatever strategy you use (network-side or MMP-side) actually works and is followed through.

Yes, CRM/owned media activities can increase the ping-pong effect. The best practice is to treat CRM efforts like any other network regarding attribution and measurement. Teams should commit to accuracy, not loosening the rules on owned media. Ensure that the same measuring stick gets applied to both paid and owned media. Implement suppression lists for users you want to reach only through owned channels and verify that these are implemented correctly by your network partners. Give your networks the lists, but continually verify that everything has been implemented correctly. Mistakes happen, so there could be some overlap. The bottom line is that anytime a suppression list is in play, a report should be in play, and you should be making sure it is actually obeyed.

This balanced approach helps prevent the internal competition that can arise when different departments optimize for their own KPIs rather than overall business outcomes.

Probabilistic attribution generally relies on shared IP addresses around timestamps. For iOS, especially after ATT, probabilistic attribution is difficult. Solutions often involve targeting pre-iOS 14.5 users or employing “funky math” to factor up based on observed Android data. Social partners such as TikTok or Meta model conversions within their walled gardens, but distinguishing observed from modeled conversions remains challenging.

Your best approach combines first-party data suppression lists, partner-specific solutions, and shorter attribution windows to minimize probabilistic “messiness.” And this raises another important question:

Exactly right. Due to ATT restrictions, audience-list-based retargeting campaigns inherently target only double opt-in users, as user identifiers are required for list syncing. For the broader opt-out audience, your options include SKAdNetwork-based campaigns that leverage aggregated conversion data and contextual targeting.

Our webinar discussion highlights an additional approach: Analyze trends in your double opt-in users and carefully extrapolate to the broader audience. Just remember that these consenting users likely represent a fundamentally different cohort than opt-out users—more engaged, tech-savvy, and potentially higher-value.

Given the challenge of expiring player contracts in your sports game, a recommendation could focus on bringing users back to an experience that emphasizes fresh content, community engagement, and personalized recommendations. Concentrating on new game modes or features can draw users back into the game, irrespective of specific game mechanics expiring (i.e., contracts). Personalizing the experience using data on past gameplay can further help provide users with more attractive reengagement. Moreover, highlighting social aspects such as leagues or tournaments also gives returning users appealing reengagement that’s less affected by expiration of player contracts (in this case).

The key is avoiding reengagement that directs users straight to a frustrating experience, possibly reinforcing their reasons for churning in the first place.

Scrutinize your data. Look at make, model, and operating system. Start by looking at post-install events. Check performance differences between attributed and unattributed users, and check differences by attribution sources. Compare internal diagnostics with MMP reporting. Breaking down user behaviors by cohort can reveal important patterns, while comparing organic and paid user engagement can highlight potential issues with specific acquisition sources. In some cases, the pattern may indicate fraud—per Simmons, “Digital ad fraud is still the second-largest criminal organized enterprise on the planet.”