Google App Campaigns (Formerly UAC): How They Work, What They Cost and How to Set Them Up

If you have searched for "UAC Google" and "Google App campaigns" and found what look like two different products, you have found the same product under two names. Google renamed Universal App Campaigns to App campaigns in 2019 without changing how they work, and the old name is still searched far more often than the new one.

The name is the least confusing part. App campaigns take away most of the controls a search advertiser relies on — no keywords, no individual ads, no placement choices — and replace them with a budget, a target, a conversion event and a set of assets. Get those four inputs right and the system has something to optimise. Get them wrong and no amount of tinkering afterwards will rescue it.

This guide covers what App campaigns are, which type and bid strategy to choose, what they cost to run properly, how to measure them on Android and iOS, and the handful of mistakes that sink most of them.

Quick answer: Google App campaigns — called Universal App Campaigns (UAC) until 2019 — promote a mobile app across Google Search, Google Play, YouTube, Discover and the Display Network from one campaign. You supply text, image and video assets, choose installs, in-app actions or pre-registration as the goal, and set a budget and target. Google builds the ads and chooses where they run. Google recommends a daily budget of at least 50 times your target cost per install.

What does UAC mean in Google Ads?

UAC stands for Universal App Campaigns, the name Google used for App campaigns until it dropped the word "Universal" in 2019. The rename, announced in February 2019, did not change features or how the campaigns work (Search Engine Journal); it simply brought the naming in line with the rest of Google Ads. Guides, agency pages and job listings that still say "UAC" are describing today's App campaigns.

How do Google App campaigns work?

You provide assets and a goal; Google builds the ads, chooses the placements and sets the bids. There are no keywords to research and no individual ads to write (About App campaigns).

One App campaign can show ads across:

• Google Search — including search partners.
• Google Play — search results, related-app sections and the Play homepage.
• YouTube.
• Discover on Google Search.
• The Google Display Network — including Gmail, other apps and mobile websites.

The app has to be published on Google Play or the Apple App Store before it can be promoted (pre-registration campaigns are the exception, covered below). From your assets, Google tests combinations and serves the ones that perform best for each network and format.

That automation is the product. It is also why the inputs matter so much: the conversion event you choose is the only definition of "a good user" the system ever receives.

Which type of App campaign should you use?

Choose by the stage of the app, not by the budget. There are three types (types of App campaigns):

For most apps the decision is simple. Before launch on Android, pre-registration. After launch, installs. Engagement comes later, once the user base is large enough to qualify and there is a specific in-app behaviour worth paying to bring people back for.

Which bid strategy should you choose?

Match the bid strategy to what a valuable user actually does in your app, and fund it properly. Google offers target-based and maximise strategies (bidding in App campaigns), and its best-practice guidance sets minimum budget multiples for each (App campaign best practices):

The most common strategic error sits in the first row. Installs are cheap to buy and easy to report. But an app whose business depends on subscriptions, bookings or purchases is not helped by an install that never opens twice, and a campaign told to find installs will find the people most likely to install — not the people most likely to pay.

How much do Google App campaigns cost?

There is no fixed price: you pay per install or action at auction, and the real cost floor is the budget Google needs to learn. Two practical points follow from Google's own guidance.

First, the budget multiple sets your minimum spend. Worked through in rupees:

• A target CPI of ₹40 implies a daily budget of at least ₹2,000 (50 × ₹40).
• A target CPA of ₹300 for a first purchase, on an install campaign, implies at least ₹3,000 a day (10 × ₹300).
• The same ₹300 target on an engagement campaign implies at least ₹4,500 a day (15 × ₹300).

These are arithmetic from Google's ratios, not benchmarks — your actual CPI and CPA depend on category, country and competition. But they show why a ₹500-a-day App campaign with a ₹300 CPA target rarely leaves the learning phase: the budget buys fewer than two conversions a day.

Second, keyword research tools overstate relevance here. Keyword Planner shows high bids for terms like "google app campaigns", but App campaigns do not bid on keywords. Those figures describe what advertisers pay for clicks on the search term, not what an install costs.

What assets does an App campaign need?

Enough variety across formats that Google can build an ad for every placement it might serve. The current limits and recommendations (App campaigns specs):

Missing formats quietly limit where the campaign can show well. Without a portrait video, for example, Google has no native creative for full-screen vertical placements, so it either adapts another asset less effectively or serves somewhere else. Google recommends an Ad Strength of at least "Good".

How do you track installs and in-app actions?

Through Google Analytics for Firebase, a third-party app attribution provider, or — for Android — Google Play directly. Installs and first opens can be measured by default; the in-app actions that matter require the relevant event to be sent from the app (mobile app conversion tracking).

On Android, Google Play reports downloads, in-app purchases and pre-registrations without extra code.

On iOS, Apple's privacy changes limit user-level measurement. Google's recommendation is the Firebase SDK, which integrates with SKAdNetwork, Apple's privacy-preserving measurement framework. Expect iOS reporting to rely more on aggregated and modelled data than Android reporting does, and plan budgets with that uncertainty in mind.

Whichever route you use, enable auto-tagging, and choose a conversion window that captures most of your conversions — Google suggests one that covers around 90% of them while staying under 30 days.

How long should an App campaign run before you change it?

Until it has recorded around 100 conversions. Google's guidance is explicit that changing a campaign before the first 100 conversions can disrupt learning, and that bids should not be cycled up and down — make one considered change rather than several small ones.

In practice that means deciding the budget and target before launch, not after the first slow week. If the budget cannot reach 100 conversions in a reasonable window at your target, the fix is to change the target or the conversion event, not to adjust daily.

Why do App campaigns fail?

The failures are rarely mysterious. The usual ones:

1. Optimising for installs when the value is in-app. Cheap installs, weak retention, no revenue.
2. Budget below Google's multiple. The campaign never gathers enough conversions to learn.
3. Changing targets before 100 conversions. Each change restarts part of the learning.
4. Missing asset formats. No portrait video or no square image means lost placements.
5. No in-app event tracking. Without it, a tCPA or ROAS strategy has nothing to optimise towards.
6. Tracking the wrong event. A "tutorial complete" event that everyone triggers is no better than an install.
7. Starting an engagement campaign too early. It requires at least 50,000 installs.
8. Judging iOS with Android expectations. iOS data is more aggregated; compare trends, not daily numbers.

What we can and cannot evidence

Being straight about this, because app marketing is full of unverifiable CPI claims.

What PPCPartner publishes: ten case studies from nine live Google Ads accounts — six Gulf, two Indian and one UK — with figures from account reporting.

Sample, market and method: nine accounts across the UAE, Saudi Arabia, Qatar, India and the UK; each study states its own period; figures come from Google Ads reporting on accounts under active management.

What is excluded: none of them is an App campaign. There is no published cost per install, retention figure or in-app ROAS in that set, and we are not going to borrow a lead-generation conversion rate and present it as app performance.

What does transfer: the parts of App campaigns that decide results are Smart Bidding against a well-chosen conversion event, correct conversion tracking, and enough budget and patience for the system to learn. Those are the same disciplines our lead-generation accounts run on — but if a published app track record is your appointment criterion, ask for it specifically.

When App campaigns are not the right tool

• Your business converts on the web, not in an app. Search and Performance Max send people to a website; App campaigns send them to a store listing.
• You cannot track any in-app event yet. Fix measurement first, or accept that you are buying installs only.
• Your budget cannot reach Google's multiple. A campaign that never exits learning costs money without producing a usable result.
• You need keyword or placement control. App campaigns do not offer it; if a specific search term matters, a Search campaign pointing to the app's landing page gives you that control.
• Your app is not yet on Google Play or the App Store — unless it is an Android app eligible for pre-registration.

What to do next

Before launching, write down four things: the in-app event that defines a valuable user, the target cost for that event, the daily budget that clears Google's multiple for it, and the date by which you expect 100 conversions. If those do not add up, change the plan before you change the campaign.

We set up and manage App campaigns as part of Google Ads management, and if the measurement layer is the gap, conversion tracking is where to start.