Studios often ask us the same thing when they plan a channel mix: Unity Ads vs Google UAC, which gives better quality for mobile game installs? The honest answer is that neither wins everywhere. They find players in different places, optimise in different ways and need different things from your data. The better question is which one fits your game, your goal and your stage.

This post compares the two fairly. We cover how each platform works, what “quality” should mean, where each tends to perform better, and how to run a test that tells you the truth. Both are part of the mix we manage, alongside Applovin and Mintegral, so we have no reason to push one over the other.

What “better quality” actually means

Quality is not a low install price, and it is not a high retention number on its own. For a game, quality means the players you buy go on to return more revenue than they cost. That is payback, and it depends on how your game earns: ads, in-app purchases or both.

So before comparing platforms, agree on the deciding metric. We use matured cohort ROAS at a fixed day, and retention as a supporting signal. Early metrics such as cost per install or installs per thousand impressions are useful to shortlist, but they cannot settle a quality question. Our guide to what a good ROAS looks like explains how to set that target for your game.

How Unity Ads user acquisition works

Unity is a network with a large footprint inside games. That matters because your ads appear to people who are already playing mobile games, often in formats such as rewarded video, interstitials and playables. If your game is a game for gamers, that context is a plus.

On the bidding side, Unity offers ROAS campaigns driven by machine learning. You pick an optimisation type that matches how you earn: in-app purchase, ad revenue or a hybrid of both. You also pick a prediction window. According to Unity’s own documentation, the windows run from day 0 for ad revenue up to day 28 for purchase and hybrid campaigns, with the longest window still marked as beta. Shorter windows adapt faster, longer ones lean towards lifetime value. You then set a target ROAS, and a lower target brings more installs while a higher one brings fewer, higher value players.

Two practical points from that documentation are worth knowing. First, Unity needs post-install event data, and an integration with a mobile measurement partner is the recommended route. Second, new ROAS campaigns start in a learning phase, and it ends only once the campaign has seen enough purchasers or enough attributed installs with matured revenue. Until then, performance can swing and the target may not be met. If you judge it in the first days, you will judge noise.

How Google Ads app campaigns work

Google Ads runs app campaigns with very little manual targeting. You supply creative assets, a budget, countries and a bidding goal, and the system assembles ads and places them across Google’s inventory, including search, the Play Store, YouTube and a network of partner apps.

That reach is the main strength. Google can find intent that no in-game network can, such as people searching for a type of game or browsing the store. It is also why Google works well for scale in many genres.

The bid strategies step up in depth. According to Google’s documentation, you can optimise for install volume, for a specific in-app action or for in-app action value with a target ROAS. A few rules from the same documentation are useful in practice. Choose one key action for action bidding rather than several, because multiple actions blend into a single target with mixed values. For target ROAS, make sure the events you target actually carry conversion value. And avoid big changes before the campaign has registered roughly a hundred conversions, because edits during learning can disrupt it.

Where each tends to perform better

These are tendencies from our own work, not guarantees, and your game may behave differently.

  • Casual and hyper-casual games with ad-driven revenue. Unity’s in-game audience and its ad revenue ROAS option suit games where most of the money comes from ad views. These players are used to watching ads for rewards.
  • Midcore and strategy games with purchases. Google often brings players with clearer intent, and the in-app action value bidding fits games where a small group of buyers drives revenue. Unity can also work here when you use a purchase or hybrid campaign with a longer prediction window.
  • Puzzle and word games. Both can scale, and the difference usually comes down to creative and country. This is a case where testing beats opinion.
  • New launches with little event data. Google tends to be easier to start with an install volume goal, and then move to action goals once events flow. Unity ROAS campaigns need data to learn, so they get better after your event setup is solid.

If your game earns from both ads and purchases, a hybrid goal on Unity and a value-based goal on Google are the closest like-for-like options. We cover that setup in more detail in our look at the best UA platforms for mobile games.

How to test them fairly

Most comparisons go wrong because the two campaigns were not set up on equal terms. A fair test needs these conditions:

  1. Same countries and same app version. A different country mix changes costs and revenue per player.
  2. Comparable creatives. Run the same core concepts on both, adapted to each format. We explain how to set that up in our creative testing framework.
  3. The same deciding metric. Use matured cohort ROAS and retention at the same day for both.
  4. Enough time. Let both pass their learning phase before you read results. Judging early favours whichever platform happens to settle first.
  5. A separate budget for the test. Do not take it from your main campaigns, so a weak result cannot hurt your core spend. Our guide to planning a UA budget shows how to hold a standing test allowance.

Also watch for overlap. Players can see ads from both platforms, so attribution from your measurement partner is the source of truth. Do not rely on each network’s own reports to decide.

Do not treat it as either or

In our experience the strongest setups use more than one network, because each has a ceiling. As you scale, installs get harder to find on any single platform, and quality can slip. Adding a second source of players often restores both. Across more than three years of aggregated results, our blended ROAS has stayed above break-even, and that comes from balancing sources, not from betting on one. Applovin leads the mix for us, followed by Google Ads, Mintegral and Unity, but the right order for your game may be different.

Our three-phase approach to scaling, covered in the growth roadmap, puts channel expansion in the right place: after your first channel is stable, not before.

A simple way to decide

If you are choosing where to start, use this order of thinking:

  • If your game earns mostly from ad views, start with Unity and an ad revenue goal, and add Google when you want more reach.
  • If your game earns mostly from purchases, start with Google on an in-app action and move to value bidding once you have enough events, then add Unity with a purchase or hybrid goal.
  • If you are not sure, split the test budget evenly, wait for matured cohorts and let the numbers decide.

Whichever route you take, keep the rules simple: one question per test, one deciding metric and enough patience for learning phases to end. Quality is something you measure over weeks, not something a platform’s name guarantees.

If you want help setting up a test like this for your own game, get in touch and we can talk through what your data allows.