The cheapest way in
is $14.99, and it
expires in six hours.

A mature live game has two honest ways to raise revenue per user: take more from the people who already pay, or turn more of the people who don't into people who do. Which one you pick determines every downstream decision, so it's worth settling first — and it settles quickly.

Extraction is out. WWE Champions carries 50M+ installs and a 4.36★ rating across 402,333 ratings. Whales on a title this age are already monetised close to their ceiling, so there is little headroom — and the review corpus prices the goodwill cost precisely: “money” is the most-used word across 2,945 reviews, and it skews +26.7 points toward one star. That leaves conversion.

And conversion is a first-week problem. In mid-core RPGs the large majority of players who ever pay make their first purchase within days of installing; the payer / non-payer split is set early and rarely re-opens. So ARPDAU on this title is decided in the same ten minutes the previous version of this case worried about — but the thing to fix in those ten minutes isn't retention. It's the shelf, and specifically the bottom of it.

The licence isn't the problem either, which is the most useful thing the competitive set tells us. WWE SuperCard is rated 4.77 by 2.6× as many iOS raters, and Empires & Puzzles launched five weeks later, same genre, no licence at all, and carries 3.4× the iOS ratings at 4.69. Same nine years, same mechanic, no IP advantage. Whatever is costing this title goodwill, it isn't the audience and it isn't the wrestling.

Evidence · lifecycle numbers, the competitive table, and what I couldn't verify →

An earlier version of this case chose “the brand-new installer.” That answer isn't wrong so much as underspecified, and the difference is the whole rewrite: retaining installs grows DAU, which is the denominator. The early-game spender is the subset of that same group where the money actually is — same ten minutes, same screens, different success metric.

The mid-game staller and the lapsed player keep their previous verdicts — real, but deferred, and for unchanged reasons. Both appear in what we're parking, each with the condition it's waiting on.

Evidence · the five pain points and 42 screenshots from my own first session →

That is not the sound of a dying game, and it is not the sound of an audience that resents paying. People who don't care about spending don't talk about money. This is high intent meeting a hostile transaction — a conversion problem, and conversion problems are cheaper to fix than desire problems.

Sorting all seven complaint themes by where in the purchase path each one stops someone is where the story gets sharper than “players hate the store”:

Keyword families over the 2,945-review corpus; a review can match several themes, so the percentages describe reach and do not sum to 100. “Avg rating” is the mean star rating of reviews matching that theme.

The two shaded rows are the first purchase — and they are the loudest and the angriest themes in the corpus. Money is the most-reached complaint at 36%. Trust is the lowest-rated at 2.06★. Both sit directly on P0→P1, and neither has ever been the subject of product work.

Evidence · all four views of the corpus, and the method behind them →

Filter one — numerator, not denominator. The decomposition at the top of this page does most of the cutting on its own. ARPPU is defended and DAU is the denominator, so anything that doesn't raise the share of players who pay has to justify itself unusually hard.

Filter two — does it compound? A single well-placed offer converts someone once. A ladder converts them and then gives them somewhere to go. Repeat-purchase rate is the compounding term in this model, which is why the recommendation is a three-rung structure rather than one starter pack — and why the second and third purchase get as much design attention as the first.

Filter three — risk to the golden goose. This game's revenue works. Anything that touches whale pricing or existing gacha drop rates can damage a functioning economy faster than a conversion gain repays it. Deterministic starter bundles are the safe surface: they add rungs below the existing floor without moving any rung already on it.

That asymmetry is why first-purchase conversion outranks the retention bets even on retention grounds, and it's where the downstream D7 and D30 effect comes from: payers retain far better than non-payers, so a conversion win is a retention win you get for free — while a retention win is not a conversion win at all. One caveat, stated plainly: that relationship is correlational until it's matched on session count, because payers were always the more engaged players. That match is data pull 3, and the downstream line is deliberately the smallest of the three revenue components in the model, so the case doesn't lean on it.

Evidence · the model, every input editable →

Base-case annualised incremental revenue at the model's assumptions, which are stated and editable on the evidence page. Move 03 scores highest despite being the cheapest because it's the only one that lifts repeat rate — and the repeat assumption is the softest input in the model, which is exactly why data pull 4 exists. Move 02, the one the other two depend on, is now the smallest of the three by value: a $14.99 pack already exists, so it deepens a ladder instead of creating one. Structurally load-bearing and financially third — both are true, and the sequencing argument below is what justifies building it anyway.

Why in this order — each necessary, none sufficient

Evidence · the mechanism and guardrail behind each move →

Both rows scored the same way against the same baseline: total DAU 121,212, ARPDAU $0.2200. Marginal retained users are assumed to monetise at half the average — see below for why, and for what happens if you don't assume it.

So take the smaller number seriously rather than hiding it. $672k of retention revenue is real money and I'd take it. But if ARPDAU is the number on the scorecard, a bigger revenue figure earned by adding non-payers is the wrong trade — it looks like progress in the board pack and reads as decline in the metric. The two jobs are different, and this case is scoped to the one that was asked for.

The programme, by move

And by scenario

A +0.80pp move on ever-pay conversion is a +40% relative lift — cut from +50% once the $14.99 starter pack turned up in my own screenshot, because two rungs beneath an existing pack are worth less than a ladder built from nothing. The premise that survives is that nothing sits below $14.99. If pull 2 finds a working sub-$10 offer already in front of new players, the conservative row becomes the ceiling and move 02 gets re-scoped again.

Evidence · open the model and change any input →

1. The payer funnel by session index. Install → session 1…7 → first store impression → first purchase. Where does a first purchase happen today, how long does it take, and at what price? This is the query the whole case is asking for.
2. Price-point distribution of first purchases, and the SKU ladder a new account can actually see. Is there anything under $10 in front of a new player? What's the modal first-purchase price? This confirms or kills move 02.
3. D7 and D30 for payers vs non-payers, matched on session count. The matching is the point — unmatched, it only tells you that engaged players spend. Sizes the downstream retention line and tests the asymmetry the whole priority order rests on.
4. Second-purchase rate and time-to-second, for first-time payers. Sizes the compounding half of the model, currently its softest input.

Then build, in this order:

Published odds ships in phase 1 rather than with the rest of move 03 because it's largely configuration and legal review — already table stakes in several markets — and it needs to be live before the rungs are, not after. Phases 1 and 2 overlap deliberately: they touch different surfaces and can be split-tested independently on the same install cohort.

Evidence · the full specs, pod-ready →