Studios Built on Roblox Did Not Get to Set Their Own Revenue This Quarter

Nine Billion Dollars in One Day

Roblox closed at $35.60 on July 31, down 26.9% (GuruFocus). About $9 billion in market value disappeared in a single session (TheStreet). It was the worst day since the company went public.

The quarter itself was not the problem. Revenue came in at $1.47 billion, up 36%. Bookings were $1.56 billion, up 8%, landing at the low end of guidance. Net loss narrowed to $185 million from $280 million a year earlier, and free cash flow rose 66% to $294 million. Daily active users reached 123 million, up 10%, and hours engaged reached 29 billion, up 5%.

What broke was the forward look. Third-quarter bookings were guided to $1.576 billion to $1.653 billion, a decline of 14% to 18% year over year. The full-year outlook was withdrawn outright. A growing company told the market it would shrink by double digits next quarter.

The reason management gave is what makes this worth reading. No external shock, no new entrant, no regulator. It was the recommendation algorithm the company itself changed in April.

What Changed Was the Scoring Rule, Not the Ranking

The April change did not reshuffle a leaderboard. It replaced the measurement underneath it. Discovery had been evaluating experiences on 7-day player retention. It now evaluates them on 28-day retention. The question moved from whether a game holds you for a week to whether it holds you for a month.

The shareholder letter is blunt about the consequence. The change “intentionally provides more impressions for highly retentive games at the expense of near term monetization.” The word intentionally sits in the sentence. This was a design decision, not a regression.

CEO David Baszucki narrowed where it landed: the US under-13 cohort, showing up as a lower frequency of impressions for games built around short-term monetization. CFO Naveen Chopra put the same thing in financial terms, describing a greater than expected shift of engagement away from “high-monetizing 2025 vintage viral games” toward “a combination of new and evergreen experiences with lower hourly monetization.”

The sequence is simple. The platform redefined what a good game is, redistributed impressions to match the new definition, and the money followed the impressions. Roblox published what that redistribution cost the parent company: 14% to 18% of a quarter’s bookings. What it cost or gained each individual studio underneath is disclosed nowhere.

Traffic Held, Revenue Did Not

The number to carry away is not the share price. It is that two metrics pointed in opposite directions in the same quarter. Users up 10%, hours up 5%, bookings guided down 14% to 18%. Nobody left. The money simply exited through a different door.

Revenue on a UGC platform is a product, not a sum. What you built, multiplied by the impressions the platform allocates you, multiplied again by the payout rate the platform sets. You own the first term. You own neither of the others. Roblox touched both this year. April rewrote impression allocation, and on June 8 the company introduced a separate DevEx rate of $0.0054 per Robux, 42% above the standard rate, applying only to spend from age-verified US players 18 and over. Roblox justified it by noting that its US 18 to 34 cohort is growing more than 50% year over year and monetizes 50% higher than the under-18 cohort.

Read that as a creator and it says something uncomfortable: the same game, run by the same developer, now pays out at different rates depending on who happens to be spending. Impressions shrink for studios extracting quickly from young players, and payouts rise for studios earning from adults. The platform is changing its own audience mix, and it is dragging the creator portfolio along with it.

The genuinely unusual part is that any of this has a number attached. App Store rankings, YouTube recommendations, Amazon search placement, and Steam’s discovery queue all perform the same function, and none of them disclose when they are retuned or what the retuning cost the businesses sitting on top. Roblox is public, so it had to write the invoice for its own reallocation into a guidance range. Studios on the platform got a figure they would normally never see. Distribution being someone else’s asset is usually an abstraction. This quarter it came priced.

What to Check Now

Split your revenue in two. Separate what came from sessions that began with a platform recommendation or search from what came from people who arrived directly because they already knew you. If the first bucket is more than half, then half your revenue is allocated rather than earned. Plenty of teams are reading an allocation curve as a performance curve and do not know it.

Then look at where the scoring rule moved. Impressions now flow to games that still have you back on day 28. If your monetization is engineered to convert in the first session, that design is currently taxing your own distribution. This is not Roblox-specific. App stores and marketplaces have been folding retention and repeat purchase into placement for years, and installs alone stopped being the currency a while ago.

Fix any revenue model that treats the payout rate as a constant. Roblox introduced an entirely new rate tier inside a single year, and it was conditional on who the payer was. Model the rate as a range instead of a point, and decide in advance what you cut at the bottom of that range.

Finally, if one platform carries most of your revenue, the quarter to test a second channel is this one, not the one after the reallocation reaches you. The market did not learn about an April change until the end of July. That lag is the whole warning.