Adikteev's Emilien Eychenne knows that your biggest game growth lever isn't new installs; it's learning to spot churn early and win mobile players back for good.

You pay good money to acquire a new player. Their activity reveals details about them, and you have every opportunity to monetize them. Unfortunately, most games let that player walk out the door without any marketing follow-up. 

I have spent more than thirteen years building mobile re-engagement and retargeting platform Adikteev, and almost as long investing in gaming, adtech, and AI companies. Running one company teaches you how to survive inside a single set of constraints. Watching twenty companies at once teaches you how many different ways there are to win. One lesson keeps coming back to me: Most teams pour their energy into the top of the funnel and accept the leak at the bottom. That leak is where the real opportunity lives.

Here is a factor most people don't know: a returning player costs less to reach, arrives with higher intent, and comes with a track record you can actually learn from. Even the smallest leak can sink a large ship over time.

Retargeting is underused because it is less glamorous than hacking growth through new installs. However, the math is in your favor once you put in real effort. Remember that you are not chasing a stranger; you are doubling down on someone whose value you have already measured.

Churn is a slope, not a cliff

The decision to leave is usually made before a player even opens your app. That's what makes churn so hard to predict in a live game: wait for the obvious signal, and you're already too late.

The signals that matter show up early and quietly: sessions get shorter, the gap between sessions stretches out, and the player starts skipping the parts of the loop they used to love. Spending slows and progress stalls. In a mobile game, the window that decides everything is often the first few days, and the moment right after a hard difficulty or content wall.

Properly monitored, you can see disengagement forming days before a player consciously quits, while there is still a live relationship to act on. The hard part is not deducing who left; it is spotting who is about to leave, early enough that a thoughtful nudge still lands.

Value and the sales funnel fallacy

Retention is one of the hardest problems in games, and for a structural reason: your product has to earn attention every single session against endless alternatives. There is no lock-in, no switching cost, and no work reason to come back. That’s why day 1, day 7, and day 30 numbers fall off a cliff. Most of those players were never going to stay. You are planning for the few who will.

The usual killers are the same across studios. Onboarding that does not deliver the core fun fast enough. Difficulty spikes that feel like punishment. Long stretches with nothing new to pull a player back.

The biggest missed opportunity is not acquiring more players; it is the wavering player in the middle of the funnel who had one bad session and is quietly drifting away. That player is still reachable, but most studios simply do not look.

Bring players back for a reason, not with a bribe

There is a real difference between reaching a player at the right moment with the right message and yanking them back in a way that does not stick. An artificial win-back, a generic push, or a bribe buys you one session and a nice bump on the dashboard. Then the same churn returns, this time with a slightly more annoyed player.

Durable re-engagement starts with why this specific person disengaged, and meets them at a moment that genuinely matters to them. At Adikteev, we spend a lot of time understanding the data before we launch anything, because a well-informed campaign beats a rushed one every time.

If you measure success by "did they open the app once," you are optimizing for spikes.

The metric you choose shapes the campaign you build. Judge yourself on downstream retention and long-term value from the players you bring back, and you are forced to reconnect them to what they enjoyed. We pilot on return on ad spend, and we are quick to pause anything that is not working, though we never judge results on day one alone. The real test is not whether a player returns; it is whether they are still there weeks, or even months, later.

What ATT changed, and what it did not

The end of easy user-level tracking hit our industry hard. App Tracking Transparency removed the stable identifier that retargeting had quietly been built on. It forced us to rethink our whole approach.

What we rebuilt was the layer underneath. We moved from following a single identified person to using privacy-safe signals, aggregated measurements, and modeling. In practice, retargeting became less about "I know exactly who this is" and more about "I can identify the right player and the right moment with high confidence." The teams that survived did not chase loopholes to recreate the old world. They rebuilt on first-party data, context, and better models. Our own answer was a proprietary ID layer that recognizes users without the old advertising identifier with about eighty percent confidence, which is more than enough to make retargeting them worthwhile.

Where AI helps, and where judgment still wins

AI is genuinely changing what is possible in prediction and personalization. It reads the subtle disengagement signals, decides who to reach and when, and tailors campaigns at a scale no human team could match. Generative tools now speed up creative work and free people from low-value manual tasks, allowing them to focus on what matters most: the relationship with the client and the player.

But AI is not ready to define strategy. It does not understand what a healthy, long-term relationship with a player actually looks like. It is great for fast, creative testing, but still short on generating the ideas that truly move people to come back.

The winning setup is AI doing the analysis and the short-term optimization, and human judgment defining what "good" means.

Re-engagement needs a predictive system

Privacy rules keep tightening, but the pace has slowed. Google has delayed removing its advertising ID, and the big platforms are under legal pressure over how they enforce these rules. I do not expect dramatic shifts on that front soon.

The bigger change is that AI keeps narrowing the gap between predicting churn and responding to it. Re-engagement is becoming real-time and personal rather than a set of separate campaigns. Players will tolerate less spam and reward genuine relevance more sharply. So the discipline moves away from retargeting as a tactic and toward retention as an always-on, predictive system that lives close to the product itself.

That shift rewards studios that treat retention as a system with signals and feedback loops, rather than a series of blasts. Gaming taught me that discipline, and it is the most portable thing I carry into every other category I work in.

If you struggle with re-engagement, it helps to have partners who quietly handle the parts that keep a game running and growing, so your team can stay focused on the fun.

That is the space Xsolla works in, with all the things a studio needs to launch, grow, and keep players coming back.

Interviews, copywriting, editing, and article publication prepared by Kirill Tokarev, Lisa Sirlin-Hall, Viktoria Steshina, and David Jagneaux.