Industry Notes

Silksong at eight months: what $20 actually bought, and what it cost

silksongteam-cherryindiemarketingpost-mortem

Hollow Knight: Silksong launched on September 4, 2025 at $20. Eight months later, the game has sold more than seven million copies for roughly $140 million in revenue excluding Game Pass, crashed Steam and the Nintendo eShop within thirty seconds of going live, and lost Game of the Year at The Game Awards to Clair Obscur: Expedition 33, a game it has since outsold. The $20 price has become the most-discussed indie pricing decision of the decade. The question with eight months of data is whether the math actually supports the headline.

The launch numbers, in hindsight

Team Cherry made two pre-launch decisions that produced the September 4 outcome. They priced the game at $20. They declined to offer pre-orders. The combination meant that every buyer who wanted day-one access had to purchase exactly at release — concentrated demand into a single window.

The window broke storefronts. Silksong hit 75,000 concurrent Steam players in its first ten minutes, 100,000 in 45 minutes, and peaked at 535,213 three hours after launch — the 18th-highest all-time concurrent on Steam, achieved by a $20 indie metroidvania. Five million players were in the game within three days, one million of them via Game Pass.

None of those numbers were guaranteed when the pricing decision was made. Team Cherry stated their reasoning in plainly modest language. Co-lead Ari Gibson said the team tries to “price the games at a reasonable level for people.” Other indie metroidvania developers interpreted the same decision more sharply — Team Cherry was “being cool and giving their game away” in a way most of them “can’t afford.”

The math on $20 vs the alternative

The headline claim is that $20 was the smartest indie pricing decision of the decade. The arithmetic asks for a counterfactual: would Silksong have made more or less at a higher price?

A $30 price — still well below the modern AAA standard, in line with Cuphead and Celeste’s later pricing tiers — would have to retain at least 67% of unit sales to match $20’s revenue (7M × $20 = $140M; 4.67M × $30 = $140M). Sticky indie audiences with deep enthusiasm for the brand frequently absorb $10 price differences without proportional unit drops; conservative estimates would put $30 Silksong at five to six million units in the same window. That’s $150M to $180M in launch-period revenue, roughly seven to forty million dollars more than the actual outcome.

By pure unit-revenue logic, $20 left money on the table. That’s the part of the story that doesn’t get told often.

What the unit math doesn’t capture

The unit math also doesn’t capture what made the eight-month outcome possible. Three things $20 bought that $30 wouldn’t have:

The crash itself was free marketing. Silksong crashing Steam was every gaming outlet’s lead headline for forty-eight hours. A $30 launch with a more diffuse demand curve produces a smoother day-one but a quieter news cycle. The cultural moment of “indie game broke the storefront” is not separable from the price.

Word of mouth scaled differently. A $20 game gets bought on impulse, gifted, lent to friends, recommended by streamers without “if you can spare the money” caveats. A $30 game in the same audience produces meaningfully fewer of those frictionless adoption moments. The compounding effect over three months — five million players became seven million, seven million became the largest engaged metroidvania audience in history — comes from velocity, not just price.

The Game Pass economics flip the math. Microsoft pays a negotiated rate for Game Pass titles, largely insulated from list price. The millions of Game Pass players Silksong picked up wouldn’t have changed at a higher list price, but the audience funnel from “I tried it on Game Pass and want to own it on Steam” benefits from a low Steam price. The dual-platform play optimises at $20 in a way it doesn’t at $30.

The honest read is that $20 wasn’t the revenue-maximising decision in pure unit math. It was the audience-and-brand-maximising decision in compounding effects. Those are different optimisations.

The Game Awards moment is its own data point

The other story Silksong’s eight months wrote is the gap between critical award recognition and commercial reality. Clair Obscur: Expedition 33 swept The Game Awards in December 2025, winning Game of the Year with twelve total nominations — the most-nominated game in the awards’ twelve-year history. Silksong didn’t win Player’s Voice either.

Six weeks later, the Steam Awards 2025 reversed the call, giving GOTY to Silksong over the same Clair Obscur. By that point Silksong had outsold every other GOTY nominee combined at The Game Awards. The Game Awards picked the auteur darling; the audience picked the metroidvania.

What that gap signals — whether The Game Awards’ voting body weighs differently from player sentiment, whether Clair Obscur’s nomination sweep reflected a particular kind of critical preference, whether Silksong’s pricing pulled it down-market in award-voter perception — is an open question. It is also, plainly, the kind of split that didn’t exist for either game when they shipped.

Replicable?

The follow-up question every indie metroidvania developer has asked publicly since September is whether $20 is a model. The honest answer is no, in the same way most things Team Cherry has done aren’t replicable: the original Hollow Knight has sold over fifteen million copies, giving the studio a balance sheet that decouples Silksong from rent-or-eat economics. A three-person indie team with a different first-game outcome can’t price the second game on long-term brand math; they have to price it for the survival of the studio.

What’s transferable is narrower. Team Cherry didn’t price for revenue maximisation. They priced for audience compounding. That distinction — and the recognition that the two are different objectives — is the part other indie teams can take from the eight-month outcome. The number $20 is not the lesson. The willingness to optimise for the wrong axis, on purpose, is.

What’s next

Sea of Sorrow, Silksong’s free expansion, is slated for 2026. Team Cherry has also confirmed a Hollow Knight: Switch 2 Edition. Both extend the audience-compounding strategy rather than monetising the existing one — a paid expansion would have been the obvious revenue play, and the studio chose to skip it. Whether the next eight months produce another commercial result the unit-revenue model can’t predict is the test that matters more than the original $20 decision.

Sources