If you last looked at mobile marketing several years ago, the mechanics have changed more than the strategy.
What happened. Device-level advertising identifiers are no longer generally available on iOS without explicit permission, and most people decline. Attribution moved to aggregated, delayed, privacy-preserving reporting. The practical effect: you can no longer see precisely which ad produced which install and what that person then spent.
What that broke. Tight per-user attribution and the fast optimisation loops built on it. Campaigns take longer to learn, small budgets learn slowly or not at all, and per-user lifetime value modelling from ad data alone is far weaker.
What people do now:
- Lean on first-party data. What happens inside your own game is still fully visible to you, and it is now the most reliable signal you have. Good in-game analytics went from useful to essential.
- Model rather than measure. Predict likely value from early behaviour instead of waiting for precise attribution.
- Contextual placement. Where an ad appears matters again, since who sees it is less knowable.
- Creative does more of the work. When targeting is blunt, the ad itself has to do the qualifying.
- Organic and community matter more, because they are not subject to any of this.
What did not change. A game with weak early retention still cannot be fixed with marketing. It was expensive to learn that before; it is slower and more expensive to learn it now, which is an argument for testing retention before spending on acquisition.
For a small studio the honest summary: paid acquisition is harder to do well at small budgets than it was, which raises the value of everything you own outright, your store page, your organic search, your community and your existing players.
Anyone running paid campaigns now, what is working?
What happened. Device-level advertising identifiers are no longer generally available on iOS without explicit permission, and most people decline. Attribution moved to aggregated, delayed, privacy-preserving reporting. The practical effect: you can no longer see precisely which ad produced which install and what that person then spent.
What that broke. Tight per-user attribution and the fast optimisation loops built on it. Campaigns take longer to learn, small budgets learn slowly or not at all, and per-user lifetime value modelling from ad data alone is far weaker.
What people do now:
- Lean on first-party data. What happens inside your own game is still fully visible to you, and it is now the most reliable signal you have. Good in-game analytics went from useful to essential.
- Model rather than measure. Predict likely value from early behaviour instead of waiting for precise attribution.
- Contextual placement. Where an ad appears matters again, since who sees it is less knowable.
- Creative does more of the work. When targeting is blunt, the ad itself has to do the qualifying.
- Organic and community matter more, because they are not subject to any of this.
What did not change. A game with weak early retention still cannot be fixed with marketing. It was expensive to learn that before; it is slower and more expensive to learn it now, which is an argument for testing retention before spending on acquisition.
For a small studio the honest summary: paid acquisition is harder to do well at small budgets than it was, which raises the value of everything you own outright, your store page, your organic search, your community and your existing players.
Anyone running paid campaigns now, what is working?