Anatomy of a launch that bought charts and lost money

Every launch season produces at least one title that spends its way into a top-ten download position, holds it for a fortnight, and is quietly unprofitable by the end of the quarter. The post-mortem usually blames the creative or the monetisation. The more accurate blame sits with a spreadsheet assumption made months earlier: that a chart position would deliver an organic multiplier large enough to make the burst pay for itself. That assumption was sound in 2019. It isn't sound in most categories in 2026, and teams are still planning launches as if it were.

The thesis, then, is that chart buying is a tactic whose underlying mechanism has weakened while its cost has risen, and that most launch plans have not been re-derived from the new mechanism. They inherited the old one.

What the chart used to do

The original logic was straightforward: store charts were a primary discovery surface, so a game in the top ten of its category reached millions of browsers who would never see an ad for it. Paid installs pushed the game up the chart, the chart delivered organic installs at zero marginal cost, and the blended cost per install fell far enough to make an otherwise uneconomic burst profitable. The paid-to-organic ratio was the whole business case.

Two things have changed at once. First, the stores have de-emphasised charts as a browse surface in favour of editorial and algorithmic placement. Apple used WWDC in June to introduce AI-driven personalised collections in the App Store, and Google Play's announcements at GDC in March were about game trials and cross-device purchasing and an AI overlay for discovery, not chart prominence. Neither company is investing in the surface that chart buying depends on. Second, the organic side of the ratio has been shrinking for years. Adjust's March report on gaming sessions and retention recorded the global paid-to-organic install ratio rising by more than half year on year, which is a polite way of saying organic is a smaller share of every install a game receives. Both changes reduce the multiplier the chart delivers per paid install.

The incentive that keeps the tactic alive

If the mechanism has weakened, the obvious question is why the tactic persists. Part of the answer is that it still works in one narrow sense: a game that spends enough does, reliably, reach the chart. The tactic delivers its visible output. What it no longer reliably delivers is the invisible output, meaning the organic lift, and because most launch plans never measured the organic lift directly, its absence doesn't register as a failure until the finance review.

The other part of the answer is that a chart position is legible to everyone. A publisher's board understands a top-ten screenshot, and so does a platform partner, and so does a press cycle. A blended cost per install that came in twenty per cent worse than plan is understood only by the UA team, and only after the fact. The incentive is to buy the thing everyone can see and hope the thing nobody measures shows up.

The second-order effect

The cost most launch coverage misses is what the burst does to the game's own auction position afterwards.

A burst campaign buys at the top of every channel's return curve. To hit a volume target inside a fixed window, the buyer accepts rising costs and widens targeting, then pushes into placements and geographies that would never clear a return threshold in steady state. The algorithms on the large networks learn from those installs. When the burst ends and the team drops to sustainable spend, the models have spent a fortnight learning from low-intent users, and the first weeks of steady-state buying are noticeably worse than they would have been without the burst. The launch does not just cost what the burst cost. It costs the recovery period as well.

There is a second, quieter effect on the game itself. A burst cohort is disproportionately made up of users who installed because the game was visible, not because it was for them. Their retention drags down the blended figures the live team uses to tune the early game, and the tuning that follows optimises for the wrong players. Teams that have seen this pattern describe the first month after a chart-bought launch as a period where every metric looks worse than soft launch predicted, and the correct explanation is that the population changed, not the game.

An illustrative reconstruction

Take a hypothetical midcore launch planned on a chart-burst model. The plan assumes that reaching a top-ten position in a mid-sized Western market will deliver organic installs equal to roughly the paid volume, halving the blended cost. The plan sizes the burst to hit the position and hold it for ten days.

The game reaches the position on schedule. Organic installs during the burst come in at a fraction of the assumed level, closer to a quarter of paid volume than a match. Blended cost per install lands well above the return threshold. The team extends the burst by a week hoping the organic curve will catch up, which it does not, because the multiplier was never going to appear. Steady-state buying after the burst runs above pre-launch benchmarks for a month as the network models recover. The quarter closes with the game short of its return target by roughly the size of the organic shortfall plus the recovery drag.

Nothing in that reconstruction involves a bad game or bad creative; it involves a plan that assumed a multiplier the market no longer provides.

A decision rule for the next launch

The alternative to a chart burst isn't spending less at launch, it's spending against a measured multiplier rather than an assumed one.

• Run the soft launch as the multiplier test. In at least one soft-launch market, spend enough to move the chart and measure the organic lift directly against a matched market where you do not. That figure, not a benchmark from a conference deck, is the multiplier the global plan should use.
• Size the burst to the measured multiplier. If the soft-launch test shows organic lift of a quarter of paid volume, the global plan should assume a quarter. If that makes the burst uneconomic, the burst is uneconomic.
• Budget the recovery. Whatever the burst costs, add a month of steady-state spend at a degraded return to the launch budget.
• Check the total. If it no longer clears the threshold, the plan was borrowing from the following quarter.

The measured multiplier is the only number in this piece that matters. Everything else follows from it, and a launch team that has one from its own soft launch is in a stronger position than one working from any published ratio, including the ones cited here. UA Ledger's earlier analysis of Supercell's invite-only approach to mo.co made a related point from the other direction: you can design a launch so that the organic side is what gets tested rather than what gets assumed.

Chart positions still have a use. They are useful for a game whose category still browses charts, in a market where organic remains a meaningful share, with a burst sized to a multiplier measured in that market. Outside those conditions, the position is a screenshot, and the money that bought it is the cost of the screenshot.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• Chart Watch, July 2025: Puzzle Holds the Download Lead
• Chart watch, March