How does a game idea pass a publisher test?

In mobile games, especially hypercasual and hybrid-casual, an idea is usually tested for a few days before anyone spends months building it. During my years at studios I ran these tests with publishers many times. This post explains how a test is set up and how the results are read.

Why so early?

Because most ideas don't work. Finding that out months later, when the game is almost finished, is expensive. A quickly built prototype and a small ad budget give the same answer for much less.

How a test is set up

1. Prototype: Only the core mechanic and a few minutes of play. No menus, no shop, no long level structure.
2. Ad videos: A few short videos showing the most appealing moment of the game. These are often tested before the game itself.
3. A small budget: The videos are shown to real players on platforms like Meta or TikTok, and the behaviour of the people who install the game is measured.

Three numbers

CPI (cost per install): What it costs to bring in one player with ads. A low CPI means the idea and the ad video catch people's interest.

Retention: How many of the people who installed the game come back the next day (D1) and the days after. This is the most honest number for whether the game is actually fun.

Playtime: Time spent in the game per session and per day. For games that earn money from ads, it is directly tied to revenue.

Every publisher has its own thresholds, and these values change with genre, country and time. So instead of looking at a single number, the three are read together. A low CPI with low retention describes an idea with a good ad but a game that doesn't hold players.

If the test fails

There's no need to give up right away. These are tried in order:

• The first seconds of the ad video, the hook
• The first 30 seconds of the game
• The core loop: the thing the player does again and again

If the numbers don't move after two or three tries, dropping the idea and moving on to the next one is the healthiest decision.

Lessons I took away

• An idea that can't be told in one sentence can't be told in an ad either.
• The ad video is part of the game; it shouldn't be left for last.
• When reading the numbers, love for the game has to be set aside.
• A fast prototype is worth more than a perfect one that arrives late.