Roblox Corporation
Rating
Hold
Hold for Long-Term Compounding
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Roblox is the dominant user-generated-content (UGC) gaming platform with two-sided creator-player network effects and a proprietary developer toolchain (Roblox Studio) that has compounded for 18+ years. The Q2 2026 problem is monetisation, not audience: DAU grew 10% YoY to 123M and hours engaged 5% to 29B, while bookings per DAU fell ~2% to $12.66 as engagement shifted away from the high-monetising 2025 viral cohort and a Recommended-for-You algorithm change deliberately traded near-term spend for retention. The moat sits in the audience and the economy around it, both of which still grew; what is unresolved is whether platform take per hour recovers.
Roblox's competitive position rests on creator-player network effects, learned interfaces (Roblox Studio + Lua/Luau scripting), and proprietary behavioral data on hundreds of millions of users across billions of play sessions:
- Two-Sided Creator-Player Network: Creators bring players (their friends, audience, communities) and players attract more creators (because of the monetisation potential), a bilateral compound running since 2008. The Q2 2026 evidence is that the content side of the flywheel is widening rather than narrowing: the top 10 experiences fell to ~20% of hours from ~30% a year earlier, while experiences outside the top 10 grew hours 25% YoY and Robux spend more than 20% YoY. That matters because the bear case for a UGC platform is dependence on a handful of viral titles, and concentration moved the other way even as bookings decelerated.
- Roblox Studio: The Learned-Interface Moat: Roblox Studio is the proprietary game development environment used by every Roblox creator. It uses Luau (Roblox's TypeScript-like Lua dialect) and a deeply integrated physics, animation, avatar, and monetisation system. Creators have invested years learning Studio; ports to Unity or Unreal are operationally costly and lose the integrated avatar/economy/discovery layer. Build — announced for Q3 2026 — puts prompt-based game creation in the mobile app for 120M+ DAUs, which cuts both ways: it widens the creator funnel to non-coders, but a platform whose own tooling makes years of Studio practice optional is trading a learned-interface moat for a distribution-and-economy one. Management calls Build early-stage with no model for its impact, so this is the open question to re-test rather than a settled downgrade.
- The Older-User Mix Shift: Roblox's route out of the kids-platform framing runs through adults, and the age-check data is the first hard read on it. By the end of Q2 2026, 57% of global DAUs had age-checked (US and Australia both above 70%). Within that checked population, over-18s were 27% of DAUs globally and roughly a third in the US, with US over-18 DAUs growing 32% YoY and hours 27%, led by 42% DAU growth in the 18–34 cohort. US over-18 users monetise more than 50% higher than under-18s, so this cohort is the mechanism that could lift blended bookings per DAU back up. It is also the segment that matters commercially: over-18s are roughly 80% of the ~$200B global gaming market, and Roblox now accounts for ~4% of it against a stated 10% goal.
- Proprietary Behavioral Data: Roblox accumulates session-level data on hundreds of millions of users across billions of play sessions per year — covering engagement patterns, social graph, in-experience purchases, and discovery flow. This dataset powers recommendation, content moderation, and creator monetisation in ways no new entrant can match without a comparable user base. AI strengthens this moat: better personalisation models compound as the data scale grows.
Ten Moats Verdict
Roblox's moat structure is led by networkEffects (creator-player flywheel compounded for 18 years), learnedInterfaces (Roblox Studio + Luau scripting), proprietaryData (session-level behavioural data), and transactionEmbedding (Robux + DevEx). No status changed at the August 2026 review, and that is the substantive finding rather than an absence of one: the Q2 2026 damage was to take per user, not to the audience or the economy around it. DAU, hours, payers and content diversity all improved while bookings decelerated, which is the signature of a monetisation problem sitting on top of an intact moat rather than a moat eroding. The two live questions are named in their pillars — whether Build converts the learned-interface moat into a distribution-and-economy one, and whether the deteriorating bookings per DAU eventually reaches the creator economy that supplies the content. Regulatory pressure has moved from prospective to priced, with roughly $54M settled across five states and more litigation outstanding, but it constrains competitors serving minors on the same terms.
Roblox Studio is the proprietary game development environment for the entire Roblox ecosystem — creators invest years learning Luau, the studio editor, and the integrated avatar/economy/discovery system. Held at strong, but with a named trigger: Build, announced in the Q2 2026 letter, puts prompt-based game creation in the mobile app for 120M+ DAUs, and a platform whose own tooling makes Studio fluency optional is converting a learned-interface moat into a distribution-and-economy one. Management describes Build as early-stage with no model for its impact, so there is no materialised evidence to downgrade on yet; re-test once Build has shipped and creator counts can be read against it. Player-side, the avatar, friend graph, and discovery feed still accumulate years of personal context that competitor platforms cannot replicate.
Roblox's economy engine (Robux + DevEx + creator payouts), physics engine, avatar system, and matchmaking infrastructure represent 18 years of accumulated proprietary technical work. Replicable in principle but operationally costly and requires equivalent platform scale to monetise.
Not applicable — Roblox is a closed UGC platform; controlling access to a public data source is not part of the business model.
Game-engine engineering and live-ops talent is broadly available across the industry; no unique research scarcity creates structural barriers vs. Unity, Epic, or large game studios. AI coding tools further reduce the differentiation that specialised platform engineers once provided.
Roblox Premium subscription bundles a monthly Robux stipend with platform features but is not a deeply differentiated bundle. The platform is more about open access than bundled tiers; bundling is a modest moat at best.
Roblox accumulates session-level behavioral data across hundreds of millions of users and billions of play sessions per year — engagement patterns, social graph, in-experience purchases, friend-network discovery. This is genuinely proprietary and compounds with platform scale; AI strengthens the moat by improving recommendation, moderation, and creator monetisation models trained on the data.
Not applicable — Roblox has no government certifications, licensure, or accreditation that creates structural switching costs. Regulatory pressure is an active cost rather than a moat, and it has now materialised in cash: five state settlements totalling roughly $54M (Nevada, Alabama, West Virginia, Mississippi, South Dakota), nine further states suing, and 100+ family claims consolidated into multidistrict litigation in the Northern District of California. The settlements mandate age verification before chat, restricted adult-to-minor contact, unencrypted communications for minors and safe-content defaults under 16. The one ambiguity worth noting is that these obligations apply to any platform serving minors at scale, so they raise the cost of entry as well as the cost of operating.
Two-sided creator-player network effects, and the Q2 2026 print is the cleanest available test of them: with monetisation deteriorating, DAU still grew 10% YoY to 123M, hours 5% to 29B and monthly unique payers 15% to 27M, while the top 10 experiences fell to ~20% of hours from ~30%. An audience that keeps compounding through a monetisation reset and a safety overhaul is the definition of a demand-side moat holding. Friend-graph stickiness is now partly constrained by settlement terms restricting adult contact with under-16s, but those terms bind competitors serving the same demographic equally.
Robux is embedded in the daily spending behaviour of tens of millions of users — monthly unique payers grew 15% YoY to 27M in Q2 2026 — and gift card distribution at retail (Walmart, Target, Apple App Store) makes it the default in-game currency in the kids/teen demographic. Creator-side, DevEx payments are deeply embedded in creators' income streams, making platform exit financially costly for top earners. Held at intact rather than strong because Roblox has shown it will curtail its own transaction rails for safety or discovery reasons: disabling cross-experience game passes measurably reduced Q2 2026 bookings, and settlement terms restrict adults transferring in-game currency to minors.
Roblox is the system of record for player avatars, friend graphs, achievement history, and creator earnings/IP. Creators' Studio projects, asset libraries, and revenue history are all platform-locked and not portable. Players' avatars, inventories, and social graphs are similarly platform-bound.
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Roblox is the dominant user-generated-content (UGC) gaming platform with two-sided creator-player network effects and a proprietary developer toolchain (Roblox Studio) that has compounded for 18+ years. The Q2 2026 problem is monetisation, not audience: DAU grew 10% YoY to 123M and hours engaged 5% to 29B, while bookings per DAU fell ~2% to $12.66 as engagement shifted away from the high-monetising 2025 viral cohort and a Recommended-for-You algorithm change deliberately traded near-term spend for retention. The moat sits in the audience and the economy around it, both of which still grew; what is unresolved is whether platform take per hour recovers.
Growth Score
Q2 2026 (reported 30 July) broke the growth thesis in a different place than the prior review expected. Bookings grew 8% YoY to $1.557B, the low end of guidance, and Roblox guided Q3 bookings to $1.576-1.653B — a decline of 14-18% YoY, the first guided drop in the company's public history — then withdrew FY2026 guidance entirely and moved to quarterly-only guidance. The prior review located the risk in age-verification driving users away; that is not what happened. DAU grew 10% YoY to 123M, hours 5% to 29B and monthly unique payers 15% to 27M. What fell was take per user: bookings per DAU declined ~2% to $12.66 and hours per DAU per day fell ~4.7% to 2.59. Management attributes the shortfall to engagement rotating out of high-monetising 2025-vintage viral titles, compounded by a deliberate Recommended-for-You change that favours retention over near-term spend, plus the disabling of cross-experience game passes. GAAP revenue still grew 36% to $1.469B, but that is flow-through from elevated 2025 bookings amortised over a 27-month payer life, so it lags rather than leads. Cash generation is turning with bookings: Q2 free cash flow was $294M, and Q3 is guided to $(60)M-$5M.
Valuation Score
At ~$36 Roblox has fallen ~75% from its 52-week high of $142 and sits just above its 52-week low of $33.88, a de-rating far larger than the deterioration in the underlying numbers. On trailing figures the stock now looks inexpensive — roughly 3.2× TTM bookings of ~$7.4B and ~15× TTM free cash flow of ~$1.64B, against ~$4.3B of net cash on a ~$24B market capitalisation. The trap is that trailing is the wrong tense: Q3 2026 bookings are guided down 14-18% and Q3 free cash flow to between $(60)M and $5M, so both denominators are about to compress. The honest read is that the price already discounts a monetisation reset and the argument is over whether it is a two-quarter comp problem or a permanent reset in platform take.
The UGC Creator Flywheel
Roblox's competitive position rests on creator-player network effects, learned interfaces (Roblox Studio + Lua/Luau scripting), and proprietary behavioral data on hundreds of millions of users across billions of play sessions:
- Two-Sided Creator-Player Network: Creators bring players (their friends, audience, communities) and players attract more creators (because of the monetisation potential), a bilateral compound running since 2008. The Q2 2026 evidence is that the content side of the flywheel is widening rather than narrowing: the top 10 experiences fell to ~20% of hours from ~30% a year earlier, while experiences outside the top 10 grew hours 25% YoY and Robux spend more than 20% YoY. That matters because the bear case for a UGC platform is dependence on a handful of viral titles, and concentration moved the other way even as bookings decelerated.
- Roblox Studio: The Learned-Interface Moat: Roblox Studio is the proprietary game development environment used by every Roblox creator. It uses Luau (Roblox's TypeScript-like Lua dialect) and a deeply integrated physics, animation, avatar, and monetisation system. Creators have invested years learning Studio; ports to Unity or Unreal are operationally costly and lose the integrated avatar/economy/discovery layer. Build — announced for Q3 2026 — puts prompt-based game creation in the mobile app for 120M+ DAUs, which cuts both ways: it widens the creator funnel to non-coders, but a platform whose own tooling makes years of Studio practice optional is trading a learned-interface moat for a distribution-and-economy one. Management calls Build early-stage with no model for its impact, so this is the open question to re-test rather than a settled downgrade.
- The Older-User Mix Shift: Roblox's route out of the kids-platform framing runs through adults, and the age-check data is the first hard read on it. By the end of Q2 2026, 57% of global DAUs had age-checked (US and Australia both above 70%). Within that checked population, over-18s were 27% of DAUs globally and roughly a third in the US, with US over-18 DAUs growing 32% YoY and hours 27%, led by 42% DAU growth in the 18–34 cohort. US over-18 users monetise more than 50% higher than under-18s, so this cohort is the mechanism that could lift blended bookings per DAU back up. It is also the segment that matters commercially: over-18s are roughly 80% of the ~$200B global gaming market, and Roblox now accounts for ~4% of it against a stated 10% goal.
- Proprietary Behavioral Data: Roblox accumulates session-level data on hundreds of millions of users across billions of play sessions per year — covering engagement patterns, social graph, in-experience purchases, and discovery flow. This dataset powers recommendation, content moderation, and creator monetisation in ways no new entrant can match without a comparable user base. AI strengthens this moat: better personalisation models compound as the data scale grows.
Ten Moats Verdict
Roblox's moat structure is led by networkEffects (creator-player flywheel compounded for 18 years), learnedInterfaces (Roblox Studio + Luau scripting), proprietaryData (session-level behavioural data), and transactionEmbedding (Robux + DevEx). No status changed at the August 2026 review, and that is the substantive finding rather than an absence of one: the Q2 2026 damage was to take per user, not to the audience or the economy around it. DAU, hours, payers and content diversity all improved while bookings decelerated, which is the signature of a monetisation problem sitting on top of an intact moat rather than a moat eroding. The two live questions are named in their pillars — whether Build converts the learned-interface moat into a distribution-and-economy one, and whether the deteriorating bookings per DAU eventually reaches the creator economy that supplies the content. Regulatory pressure has moved from prospective to priced, with roughly $54M settled across five states and more litigation outstanding, but it constrains competitors serving minors on the same terms.
Roblox Studio is the proprietary game development environment for the entire Roblox ecosystem — creators invest years learning Luau, the studio editor, and the integrated avatar/economy/discovery system. Held at strong, but with a named trigger: Build, announced in the Q2 2026 letter, puts prompt-based game creation in the mobile app for 120M+ DAUs, and a platform whose own tooling makes Studio fluency optional is converting a learned-interface moat into a distribution-and-economy one. Management describes Build as early-stage with no model for its impact, so there is no materialised evidence to downgrade on yet; re-test once Build has shipped and creator counts can be read against it. Player-side, the avatar, friend graph, and discovery feed still accumulate years of personal context that competitor platforms cannot replicate.
Roblox's economy engine (Robux + DevEx + creator payouts), physics engine, avatar system, and matchmaking infrastructure represent 18 years of accumulated proprietary technical work. Replicable in principle but operationally costly and requires equivalent platform scale to monetise.
Not applicable — Roblox is a closed UGC platform; controlling access to a public data source is not part of the business model.
Game-engine engineering and live-ops talent is broadly available across the industry; no unique research scarcity creates structural barriers vs. Unity, Epic, or large game studios. AI coding tools further reduce the differentiation that specialised platform engineers once provided.
Roblox Premium subscription bundles a monthly Robux stipend with platform features but is not a deeply differentiated bundle. The platform is more about open access than bundled tiers; bundling is a modest moat at best.
Roblox accumulates session-level behavioral data across hundreds of millions of users and billions of play sessions per year — engagement patterns, social graph, in-experience purchases, friend-network discovery. This is genuinely proprietary and compounds with platform scale; AI strengthens the moat by improving recommendation, moderation, and creator monetisation models trained on the data.
Not applicable — Roblox has no government certifications, licensure, or accreditation that creates structural switching costs. Regulatory pressure is an active cost rather than a moat, and it has now materialised in cash: five state settlements totalling roughly $54M (Nevada, Alabama, West Virginia, Mississippi, South Dakota), nine further states suing, and 100+ family claims consolidated into multidistrict litigation in the Northern District of California. The settlements mandate age verification before chat, restricted adult-to-minor contact, unencrypted communications for minors and safe-content defaults under 16. The one ambiguity worth noting is that these obligations apply to any platform serving minors at scale, so they raise the cost of entry as well as the cost of operating.
Two-sided creator-player network effects, and the Q2 2026 print is the cleanest available test of them: with monetisation deteriorating, DAU still grew 10% YoY to 123M, hours 5% to 29B and monthly unique payers 15% to 27M, while the top 10 experiences fell to ~20% of hours from ~30%. An audience that keeps compounding through a monetisation reset and a safety overhaul is the definition of a demand-side moat holding. Friend-graph stickiness is now partly constrained by settlement terms restricting adult contact with under-16s, but those terms bind competitors serving the same demographic equally.
Robux is embedded in the daily spending behaviour of tens of millions of users — monthly unique payers grew 15% YoY to 27M in Q2 2026 — and gift card distribution at retail (Walmart, Target, Apple App Store) makes it the default in-game currency in the kids/teen demographic. Creator-side, DevEx payments are deeply embedded in creators' income streams, making platform exit financially costly for top earners. Held at intact rather than strong because Roblox has shown it will curtail its own transaction rails for safety or discovery reasons: disabling cross-experience game passes measurably reduced Q2 2026 bookings, and settlement terms restrict adults transferring in-game currency to minors.
Roblox is the system of record for player avatars, friend graphs, achievement history, and creator earnings/IP. Creators' Studio projects, asset libraries, and revenue history are all platform-locked and not portable. Players' avatars, inventories, and social graphs are similarly platform-bound.
Growth Analysis
Growth Drivers
Key Risk
If the monetisation reset proves structural rather than a comp-and-algorithm artifact — FY2026 bookings finishing below FY2025's $6.79B and FY2027 failing to regain it, with bookings per DAU still under $12.66 by mid-2027 — then the Recommended-for-You trade has not bought retention, it has repriced platform take, and the correct multiple is a mature-platform one rather than a compounder's. The unresolved part is duration, not direction: management declined to say whether 2026 bookings will exceed 2025 and gave no full-year frame to hold them to.
Score Derivation
78.6 base − 2.0 trajectory − 4 margin − 5 risk = 68
Base 79 (10-18% CAGR, midpoint 14%) - 2 trajectory (three of four drivers decelerating: DAU, bookings, bookings per DAU; only the US over-18 cohort accelerating) - 4 margin compression (Q3 2026 guides Adjusted EBITDA of $0-41M against $152M in Q2 2026, with explicit fixed-cost deleverage on declining bookings) - 5 keyRisk moderate = 68. The cut from the prior 79 is carried almost entirely in the CAGR base rather than in keyRisk, because the deceleration is observed rather than prospective — the guided Q3 bookings decline is a reported number, and the risk term caps at -15 and would understate it.
Price Scenarios (12–24 Months)
Valuation Multiples
| Trailing P/E (GAAP) | n/m |
| Forward P/E (NTM) | n/m |
| Price / Sales (TTM) | ~4.2× |
| Price / Bookings (TTM) | ~3.2× |
| EV / Bookings (TTM) | ~2.6× |
| Price / FCF (TTM) | ~15× |
Bookings and free cash flow are the right denominators given GAAP unprofitability, and on trailing numbers both look cheap: ~3.2× bookings, ~2.6× EV/bookings, ~15× FCF. Every one of those multiples is measured against a trailing period inflated by the 2025 viral cycle, and the company has guided the next quarter down on both. Sizing the re-rating therefore turns on FY2027 rather than the current print — a return to low-double-digit bookings growth supports a mid-3× forward multiple, while a second year of declining bookings argues for the ~2× a no-growth platform earns.
Approximate figures as of August 2026.
Where We Are vs Targets
Loading live price…
The decline in take per user proves structural rather than a comp-and-algorithm artifact, bookings fall for a second year, and Roblox is re-rated as a mature platform carrying a permanent safety cost base.
- FY2026 bookings finish below FY2025's $6.79B and FY2027 fails to regain it, confirming that the Q3 guide of -14% to -18% marked a reset in platform take rather than a comparison against the 2025 viral cycle
- The Recommended-for-You trade does not pay back: hours per DAU per day stay below the 2.59 of Q2 2026 while bookings per DAU stays under $12.66, so the retention gained never converts into spend
- Litigation escalates beyond the ~$54M already settled across five states — nine further states are suing and 100+ family claims are consolidated in the Northern District of California — with mandated design changes that permanently raise the cost of serving minors
- Multiple compresses to ~2× forward bookings, the level a no-growth platform earns, leaving roughly $13B of enterprise value plus ~$4.3B net cash
Monetisation stabilises through 2027 as the 2025 comparisons annualise out and the algorithm change matures, bookings resume low-double-digit growth on a roughly flat FY2026, and free cash flow recovers after the Q3 2026 trough.
- Bookings per DAU returns to about $12.66 or better during H1 2027 as the shift toward evergreen titles finishes working through and the cross-experience game-pass change annualises
- DAU compounds near the 10% and payers near the 15% recorded in Q2 2026, so a roughly flat FY2026 of ~$6.7-6.9B in bookings resumes low-double-digit growth into FY2027
- Free cash flow recovers to $1.2-1.4B in FY2027 after the back-half-loaded 2026 capital expenditure cycle and the guided Q3 2026 trough of $(60)M-$5M
- About 3.5× forward bookings on ~$7.5B for FY2027, plus ~$4.3B net cash
The retention trade is confirmed to have raised lifetime value, Build converts prompt-based creation into measurable content supply, and the older-user cohort lifts blended take per user above where it started.
- Build and Moments turn prompt-based creation into a measurable expansion of content supply, lifting hours per DAU per day back above the 2.72 recorded in Q2 2025 and validating the retention-over-monetisation trade
- The over-18 cohort scales enough to lift blended bookings per DAU above $14, extrapolating from US over-18 DAUs growing 32% YoY and monetising more than 50% above under-18s
- Bookings re-accelerate to high-teens growth in FY2027 toward ~$8B as Roblox moves from ~4% of the ~$200B global gaming market toward 6%, against a stated 10% goal
- Multiple re-rates to ~5× forward bookings as two consecutive quarters of recovering take per user establish that 2026 was a transition rather than a ceiling