“What is a good ROAS?” is one of the first questions studios ask us, and the honest answer is that there is no single number. A ROAS that makes one game rich will sink another. What matters is which ROAS you are looking at, over what window, and where your own break-even line sits.

This post covers the three things we check before calling any ROAS good: blended versus cohort numbers, the break-even maths, and the time window you judge it over.

Start with the formula

ROAS is revenue divided by ad spend. If a campaign spends 1,000 and the players it brought in earn 1,200, ROAS is 120 percent.

The simple formula hides two choices that change the answer completely:

  • Which revenue counts. Only revenue you can attribute to the spend, net of store fees and ad mediation share, should be in the numerator.
  • Which players and which days. Revenue from everyone in the game this month is a different number from revenue earned by the players who installed on one day, measured a set number of days later.

Those two choices give you the two kinds of ROAS that matter in mobile games.

Blended ROAS: the business view

Blended ROAS takes all revenue in a period and divides it by all ad spend in the same period. It mixes paid and organic players, old and new cohorts, and every network together.

It is the number that tells you whether the business is working. If blended ROAS sits above break-even, user acquisition is paying for itself, including the organic lift that paid installs tend to bring with them through store ranking and word of mouth.

Its weakness is that it cannot tell you which campaign did the work. A long-running game with a big base of older players can show a healthy blended ROAS while this week’s campaigns quietly lose money.

Cohort ROAS: the campaign view

Cohort ROAS follows a group of players who installed on the same day, from the same campaign, and measures what they earned by day N. D7 ROAS is revenue by day seven divided by the spend that bought them. D30, D60 and later windows extend the same group.

This is the number you use to make decisions about a campaign, a country or a creative. D7 is the most common checkpoint because it is early enough to act on and still carries a useful signal.

The trap is reading D7 as the final answer. A cohort that looks weak at D7 can break even by D60 if players keep watching ads and buying. A cohort that looks strong at D7 can flatten out if early spenders churn. Cut too fast on D7 alone and you can switch off campaigns that would have paid back, while scaling ones that will not.

We use both numbers together. Cohort ROAS tells us where to move budget. Blended ROAS tells us whether the whole system is profitable.

Work out your own break-even line

A good ROAS is one that clears your break-even point within a payback window you can afford. That point depends on your game, not on an industry average.

A simple way to set it:

  1. Decide your payback window. How long can you wait to get the spend back? Ad-driven games usually earn most of their value quickly, so a shorter window like D60 is common. Games that rely on in-app purchases often need longer.
  2. Look at your LTV curve. From past cohorts, find what share of a player’s eventual revenue typically arrives by D7, and what share by the end of your payback window.
  3. Work backwards to a D7 target. If your cohorts historically earn about a third of their D60 revenue by D7, and you need 100 percent ROAS at D60 to break even, your D7 target sits around 33 percent. Anything above that is on track. Anything below needs a reason.
  4. Add your margin. Break-even means you got your money back, not that you made any. If you want profit after costs, set the D60 goal above 100 percent and recalculate the D7 target.

This is why two studios can look at the same D7 number and reach opposite conclusions. A game whose players pay back slowly needs a lower early number to be on track than one whose players earn everything in the first week.

For context, Unity’s ROAS guide lists observed D7 medians from third-party sources of roughly 7 to 8 percent for casual games and 8 to 12 percent for hybrid-casual. Treat figures like these as a sense check, not a target. Your own LTV curve is the only reliable basis for a goal.

Hybrid games need a blended revenue view

Revenue mix changes how you read ROAS. Across the games we manage, revenue runs at about 80 percent ads and 20 percent in-app purchases. If we measured ROAS on purchases alone, almost every campaign would look like a failure. If we measured on ads alone, we would undervalue the players who pay.

So the revenue side of the formula must include both, passed back from your mediation and measurement setup to each network. Getting this right is mostly a tracking job, and it is the first thing we check in any account. Our post on monetisation design covers how to shape that mix, and our guide to scaling Applovin campaigns explains why picking the optimisation goal that matches your revenue matters so much at scale.

Why quarterly averages are steadier than monthly

Monthly ROAS swings for reasons that have little to do with your campaigns:

  • Seasonality. Ad prices rise in the busy end-of-year period and fall in quieter months, so the same campaign can look worse in one month and better in the next.
  • Cohort timing. Players acquired late in a month have not had time to earn yet, which drags that month down and flatters the next.
  • Creative cycles. A new winning creative can lift a month, and fatigue can drag the following one, even when the overall trend is flat.
  • Store and network changes. Algorithm updates, featuring and policy changes create short spikes and dips.

Looking at a rolling quarter smooths out these effects and shows the real trend. A strong month can hide a weak quarter, and a weak month can hide a strong one. We still review numbers weekly to make decisions, but we judge whether the strategy is working over quarters.

That habit is a big part of why our blended ROAS has stayed above break-even across 40 months of aggregated results. It was not one great month. It was a long run of not overreacting to single ones.

Signs your ROAS is healthy

Rather than chasing a headline number, we look for a few signals together:

  • Blended ROAS sits above break-even on a rolling quarterly view.
  • New cohorts are tracking at or above the D7 target you derived from your own LTV curve.
  • Later windows such as D30 and D60 confirm what D7 predicted, without large gaps.
  • Spend can rise in steps without ROAS dropping below the line.
  • No single network or creative is carrying the whole result.

If most of these hold, your ROAS is good, whatever the exact figure is. If they do not, the fix is usually in tracking, targets or creatives rather than in the network itself.

Where ROAS fits in the bigger plan

ROAS targets change as a game moves through its life. During soft launch you are learning what the LTV curve looks like, so early targets are rough. As you scale, the targets tighten and the payback window matters more. Our three-phase growth roadmap shows how break-even ROAS is the gate between validating a game and spending at scale, and our UA strategy and process guide covers how we set KPIs like these at the start of an engagement.

The short version

  • Use blended ROAS to judge the business and cohort ROAS to judge campaigns.
  • Set your D7 target from your own LTV curve and payback window, not from benchmarks.
  • Count ad revenue and purchases together if your game earns from both.
  • Judge strategy over rolling quarters, and make tactical moves weekly.

If you want a second pair of eyes on your ROAS targets and how they are measured, book a free call and we will walk through your numbers with you.