Getting the install is the cheap half of the problem. The expensive half is getting the app used: in most apps, a large share of people who install never open it again. This lesson covers the machinery for winning those people back — and for not losing the ones who already have you installed.
Deep links: the prerequisite infrastructure
A deep link takes a user to a specific screen in your app (a product page, a category, the cart) instead of the home screen or your website. Without deep links there is no meaningful re-engagement: sending someone to the home screen to hunt for what you promised is the best way to lose them.
Setting them up is development work (domain verification, per-platform link schemes) and then declared in Google Ads. It is the technical requirement that most often blocks everything else — worth asking your team for well in advance.
Engagement campaigns (ACe)
They aim to get people who already have the app to do something specific: come back, buy, renew. Requirements and nuances:
- They need a meaningful installed base (the public reference is around 50,000 installs) and deep links configured.
- They optimise for in-app actions, not installs.
- They go in a separate campaign from acquisition: otherwise you are paying to reactivate your own users and counting them as new ones.
- The audience is defined with your app user lists (who hasn't opened in X days, who purchased once, who abandoned a cart), just as you would with web remarketing.
The web-to-app connection
Many companies have both a website and an app, and there a silent leak occurs: someone who already has your app installed clicks your search or shopping ad… and lands on your mobile website, an environment where they have to sign in again and where they convert worse.
Google's answer is the web-to-app connection: a set of features that detects whether the user has the app and takes them straight into it via the matching deep link, from campaigns that are not app campaigns (Search, Shopping, Performance Max and, per the most recent documentation, YouTube and Demand Gen too). Google publishes substantial conversion improvements from sending users into the app rather than the mobile web — its reference figure is on the order of 2.8× — and there are A/B tests to verify it in your own case before taking the number at face value.
For a campaign manager this means two things: (1) if your company has an app, this connection is probably the cheapest conversion improvement available; (2) the real work is deep links and measurement, not campaigns.
Pre-registration: the launch case
If the app isn't publicly available on Google Play yet, pre-registration campaigns let you build a waiting list ahead of launch, so you arrive at day one with installs guaranteed. It is a campaign type with a short life and its own goal: don't mix it with the rest.
💡 Ninja trick: before spending on winning back inactive users, check why they left: if the first experience is bad, re-engagement is paying them to confirm their decision. Winning back works when the product already retains; if it doesn't, fix the product — it's cheaper.
⚠️ Pitfall: judging an engagement campaign with acquisition metrics. There are no installs to count here: it is judged by actions recovered and revenue, and its real value is incremental — how much of that wouldn't have happened anyway.
What you should remember
- Without deep links there is no re-engagement and no web-to-app.
- Engagement campaigns go separately and are judged by actions and revenue, not installs.
- If you have both web and app, the web-to-app connection is usually the most profitable conversion improvement available.
- Pre-registration is for launches and has its own goal.