When Should You Scale Your Mobile Game?

The hardest question in mobile game development is not how to build a great game. It is knowing when your soft launch numbers are good enough to scale UA spend.

Scale too early and you burn budget on a game that will not be profitable. Wait too long and a competitor reaches your audience first. The decision window is usually a few weeks and it costs thousands of dollars either way.

This guide walks through the structured framework practitioners use to make this decision.

The Three Numbers That Matter

Every soft launch decision comes down to three metrics. All three need to clear the threshold for your specific genre at the same time. One weak metric means you iterate, not scale.

Step 1 — Compare Against Your Genre Benchmark

The most important word in soft launch is genre. A D1 of 35% is strong for RPG and below threshold for Puzzle. A CPI of $2.00 is fine for Casual and too high for Hyper-Casual.

Never compare your numbers to a generic industry average. Always compare against the benchmark for your specific genre.

Step 2 — Check the D7/D1 Ratio

Once you have your D7 number, divide it by D1. This ratio tells you where your retention problem is — which is very different information from knowing retention is low.

Below 0.30 means the problem is in your first session. Players opened Day 1 but did not return by Day 7. Fix the end of your Day 1 experience before addressing anything else.

0.30 to 0.44 means the problem is in your core loop. Players came back early but dropped off before Day 7. Days 2 to 5 engagement is weak.

0.45 to 0.55 is normal decay. Check your D7 absolute value against your genre benchmark.

Above 0.55 is strong. Retention is not your constraint — focus on CPI and creative.

Step 3 — Apply the Two-Consecutive-Week Rule

Even if all three metrics clear threshold in a given week, do not scale immediately. Wait for two consecutive weeks above threshold before increasing UA spend.

Soft launch cohorts are small — typically 50 to 500 installs. Small cohorts have high variance. A single good week could be one strong creative, one strong placement, or one cohort that happened to convert unusually well. Two consecutive weeks above threshold means the numbers held across different conditions. That is a more reliable signal.

The Three Possible Verdicts

Common Mistakes That Lead to Wrong Decisions

Scaling on a single good week

One week above threshold is variance. Two consecutive weeks is signal. Studios that scale after one good week often find the numbers revert in week two at higher spend.

Comparing to a generic CPI average

A $2.50 CPI is fine for Idle/Tycoon. The same number for Hyper-Casual means UA is not viable. Always use genre-specific benchmarks.

Fixing the wrong metric

D7 is low but D7/D1 ratio is 0.42. That is a core loop problem, not a first session problem. Studios that rebuild their onboarding when the core loop is the real issue waste months iterating the wrong thing.

Ignoring Hook CTR

A low Hook CTR with a high CPI is a creative problem. You can fix it in days by testing new ad creatives. Studios that assume a high CPI means the game is not working often kill games that just had poor creative.