Vector Wins the Quarter and Loses the Tape: A Beat-the-High-End Print Meets a Flat Guide, a Withheld Outlook, and the Genie Overhang — Downgrading Unity to Hold

Results vs. Consensus

A quarter ago we upgraded Unity to Outperform on the proof we had set as our condition: a clean total-Grow acceleration, visible non-Vector stabilization, margin expansion, record cash, and a raised guide. We named the one unresolved pillar explicitly. Valuation was the binding constraint, and the only reason the rating was Outperform rather than higher conviction. This quarter the operational case extended in every direction we wanted, and the valuation pillar that we flagged as the live risk is exactly what broke, twice over: first through a Genie-driven re-rating that took the multiple apart before the print, and then through a flat guide and a withheld outlook that gave the de-rating a fundamental hook. The result is the rare quarter where the company beat its own high end on every line and the stock fell 26%. Resolving that paradox is the substance of this note.

Q4 2025 Scorecard

Year-Over-Year Comparison

The year-over-year picture is the strongest in the coverage and a clean continuation of the trend that began at Q2. Reported total revenue accelerated to +10.1%, up from +5.4% in Q3 and roughly flat-to-down two quarters before that. Crucially, the acceleration is now visible on both axes and across both segments: Grow at +11% year-over-year is a faster year-over-year rate than the +6% Q3 print, and Create at +8% reported (+16% ex-nonstrategic) is, per management, the fastest year-over-year Create growth in more than two years. The CFO framed full-year 2025 as the year organic year-over-year growth accelerated in each of Create and Grow and in Unity in aggregate, in every quarter. On the print, the durability question the bull case needed answered kept answering itself.

Quarter-Over-Quarter Comparison

The sequential move is healthy and led by both segments. Grow added roughly $20M of revenue quarter-over-quarter on continued Vector strength (Vector reached 56% of Grow revenue, up from 49% just two quarters ago), and the CFO noted Vector added more incremental dollars in Q4 than in any prior quarter, even as the IronSource Ad Network subtracted a further $7M sequentially. Create added roughly $13M sequentially, a genuine step up rather than the optical Q3 dip, as the perpetual-license noise is now fully lapped on both sides. The 200-basis-point sequential margin expansion to 25% is the operating-leverage thesis printing again, and it was struck through the Unite user conference in Barcelona, elevated sales commissions and annual bonus accruals, and increased cloud and AI-hiring spend, all of which management called out as Q4 headwinds it absorbed.

Quality of Beat

Revenue: The $503.1M is operationally driven and clean. There is no FX windfall (Unity remains predominantly USD-billed), no acquisition contribution (IronSource is long past its lapping window and now shrinking), and no one-time license element flattering the print. The upside was entirely Vector, which management said is its highest-quality revenue source, and the disclosure that January was Vector's best revenue month ever, larger than the December holiday record and 72% above January 2024, is the single most bullish operational datapoint in the release. Management reiterated the expectation that Vector exits 2026 at a quarterly revenue run-rate comfortably above $1B annualized, which would roughly double the segment's current Vector contribution. The quality of the revenue is not in question; the issue is entirely the forward guide that sits on top of it.

Margins: The 25% adjusted EBITDA margin is the structural story extended and, at the highest level in two years, ahead of even the ~23% the company had guided as roughly stable. Management attributed it to the same operating-leverage mechanics as prior quarters, accelerating organic growth at high contribution margins, while explicitly flagging that the print absorbed the Unite-Barcelona spend, the commission and bonus accruals, and elevated R&D from cloud and AI hiring. Below the EBITDA line, the GAAP loss narrowed to $(89M) from $(123M) a year ago, and the full-year story is the cleaner one: stock-based compensation fell 19% in absolute dollars and from 33% to 21% of revenue, the single most important driver of the path toward GAAP profitability. The adjusted-to-GAAP bridge is narrowing on the line that matters, which is the direction the bull case requires.

EPS: The adjusted EPS of $0.24 against a ~$0.20–0.21 consensus is a clean beat flowing directly from the EBITDA upside, not a tax or share-count quirk. The full-year cash story is what underpins it: free cash flow grew 41% to just over $400M at a 99% EBITDA-to-FCF conversion rate, with free cash flow margins expanding 600 basis points more than adjusted EBITDA margins, the cleanest demonstration yet that this model scales cash flow faster than revenue. The cash inflection we flagged as a run-rate at Q3 is now a full-year fact, and it is the strongest single argument for owning the equity at the post-print level.

Segment Performance

Unity reports two segments. Grow is the advertising and monetization business (the Unity Ad Network powered by Vector, mediation, and the legacy IronSource-era ad products), and Create is the engine-subscription business (Unity Pro / Enterprise plus the non-gaming "industry" vertical). The story this quarter is that both segments grew on both axes again, with the composition inside Grow now openly characterized by management as a quality upgrade (Vector displacing commoditized IronSource revenue) rather than the tale of two trajectories it was three quarters ago.

Grow: Vector at 56%, IronSource Approaching Immateriality

Grow revenue of $338M was up 6% sequentially and 11% year-over-year, with the entire upside-to-guide driven by Vector's third consecutive mid-teens sequential quarter. The composition shift is now the headline: Vector reached 56% of Grow revenue, up from 49% two quarters prior, and the CFO disclosed that Vector added more incremental dollars in Q4 than in any prior quarter. Importantly, management's internal analysis attributes Vector's strength almost entirely to incremental advertiser demand and improved conversion rather than to a shift of spend off IronSource, which directly answers the cannibalization question that anchored the Q2 debate. The IronSource Ad Network, by contrast, fell another $7M sequentially to 11% of Grow revenue and is guided below 6% of total company revenue in Q1, a wind-down management now frames as a quality upgrade: displacing "commoditized, lower-margin ad network revenue for deeply differentiated AI platform revenue."

"Vector experienced yet another quarter of mid-teens sequential growth, its third in a row… Vector added more incremental dollars in the fourth quarter than in any prior quarter. January was Vector's best month ever… our internal analysis shows that Vector's ongoing strength is almost entirely coming from incremental advertiser demand and improved conversion performance rather than a shift over from customers who have been reducing spend with IronSource." — Jarrod Yahes, CFO

The catch is the Q1 guide. Despite Vector guided to +10% sequentially and management noting that, excluding IronSource, Grow grew double digits sequentially in Q4, the total Grow segment is guided flat sequentially in Q1, attributed to holiday-rich Q4 seasonality, two fewer calendar days, and the continued IronSource decline. Grow is guided to return to sequential growth in Q2. So the engine is accelerating, the legacy drag is shrinking toward irrelevance, and yet for one quarter the two effects net to flat, which is a materially less exciting headline than +6% sequential and was clearly not what a beat-and-raise-conditioned Street wanted to hear.

Assessment: The underlying Grow story is strong and arguably stronger than Q3: Vector accelerating, the cannibalization concern explicitly retired by management's incremental-demand analysis, and a credible bridge to a $1B-plus Vector run-rate by year-end. The problem is entirely optical and entirely Q1. A flat sequential guide is the mechanical sum of a still-shrinking IronSource and a seasonally-softer Q4 base, not a Vector deceleration, but it removes the visible beat-the-guide cushion that did so much work in the stock the last two quarters, and it arrives the same morning as an unquantifiable narrative threat. The signal in the segment is positive; the optics of the segment guide are the opposite, and this quarter the optics won.

Create: Fastest Growth in Two Years, China the Standout

Create revenue of $165M was up 8% reported and, after lapping the $10M of non-strategic Q4 2024 revenue, up 16% year-over-year on the underlying business, which management called the fastest year-over-year Create growth in more than two years. The drivers are the same three that have compounded all year, now operating without the perpetual-license noise that flattered and then depressed prior comparisons: subscription strength as contract renewals and price increases roll through (a 5% Pro/Enterprise increase took effect January 12), continued China momentum, and breadth across the platform. The China figure is the standout: management disclosed Create grew nearly 50% in China over the course of 2025, driven by Unity's interoperability with local platforms like OpenHarmony and compatibility with consumer channels like WeChat.

"Excluding the impact of non-strategic revenue, our Create business grew an extremely healthy 16% year-over-year, powered by strength in our subscription business… Over the course of the year, our Create business is up nearly 50% in China, the world's largest video game market." — Jarrod Yahes, CFO

The Q1 framing for Create is double-digit year-over-year growth excluding non-strategic revenue, with a similar cadence expected through 2026 (excluding roughly $40M of non-strategic revenue and one-time items running off over the year). This is the same disciplined characterization that has held all year, and it sits oddly against the central bear narrative: the segment supposedly most exposed to AI-creation disruption is the one accelerating to its fastest growth in years.

Assessment: Create remains the underappreciated leg, and this quarter it strengthened on the very dimension the Genie narrative attacks. A franchise growing 16% on its strategic base, up nearly 50% in China, with a fresh price increase in effect and Unity 6 adopted faster than any release in the company's history, does not look like a business being disrupted in real time. The +16% ex-nonstrategic rate is now the clean number, durable and ARPU-led. We continue to underwrite Create at a low-double-digit-plus strategic growth rate, and note the irony that the segment carrying the most narrative risk is the one whose printed trajectory most contradicts it.

Geographic Note: China Now Roughly a Fifth of Revenue, Compounding

China continued the structural step-up we moved into the baseline at Q3, when it rose from 15% to 20% of revenue over the trailing year. This quarter management quantified the Create-specific contribution (up nearly 50% in China for the year) and reiterated the breadth of the position: Unity is fully compatible with the local platform ecosystem, holds deep developer relationships, and benefits on the Grow side as Chinese publishers use Vector for global user acquisition. Management also flagged its industry (non-gaming) business as particularly strong in Asia, citing the auto sector's use of Unity for in-dash displays.

"This growth is truly global in nature. Over the course of the year, our Create business is up nearly 50% in China… driven by our unique interoperability with local operating platforms like OpenHarmony and compatibility with popular consumer channels like WeChat." — Matthew Bromberg, CEO

Assessment: The China contribution is now a confirmed, compounding structural driver across both segments rather than the optionality we treated it as two quarters ago, and at roughly a fifth of revenue it is a genuine positive. We continue to carry the concentration and geopolitical tail as a named risk that grows with the exposure, and we note that a business this dependent on the China ecosystem is one more reason the equity should trade with a wider risk premium than a purely domestic software peer. On a 12-month view the durability is established and the read is net favorable, with the tail unchanged.

Key Operating Metrics

Key Topics & Management Commentary

Overall Management Tone: The tone shifted markedly from the assured, confirmation-of-an-inflection posture of Q3 to a defensive, thesis-defending posture this quarter, opened by a CEO who framed the call around tuning out a "noisy" environment and offering "context for why we've never been more excited," language that is itself a tell that the prepared remarks were written in response to a falling stock and a competitive narrative rather than to a clean tape. Management was most confident and most quantitative on Vector and the cash story, where the numbers do the work, and most rhetorical and least falsifiable on the Genie and world-model questions, where the strongest available answer is a logical argument rather than a metric. The single most-pressed topic in Q&A was not the guide but Genie, and the CEO's repeated framing that the time spent answering it was "really nothing to do with Genie" but rather an opportunity to explain Unity's depth of value was an attempt to convert a defensive question into an offensive talking point.

1. Google Genie and World Models: The Narrative That Halved the Multiple

The defining topic of the quarter is one Unity did not generate and cannot control: Google's Genie world-model reveal in late January, which crystallized an investor narrative that generative world models, AI systems that produce playable, interactive video from a prompt, could displace traditional game engines and, by extension, the Create franchise. The narrative took the stock down roughly 23% in a single session before earnings, and management made rebutting it the centerpiece of the call. The CEO's argument was structural: world models are a source of inspiration and assets, not a replacement for the engine, which provides the physics, game logic, networking, monetization, and live-operations systems that turn assets into a shippable, durable game.

"We believe world models are going to be a source of inspiration and assets for creators, but that they are not in any way going to replace game engines. They are complementary, not duplicative. The kind of video-based generation that world models are good at is exactly the type of input our AI workflows are designed to leverage… Unity is not an interactive video generator, it's a 3D execution platform designed to build once and then run everywhere." — Matthew Bromberg, CEO

The CEO went further on the offensive, arguing the broader AI-creation wave is a tailwind: it will lower the barrier to creation, draw "tens of millions" of new creators into interactive content, multiply the volume of content, and thereby make discovery (Vector's job) more valuable, not less. He repeatedly cautioned investors against "overreacting to LinkedIn posts," a pointed reference to the social-media origin of much of the recent competitive anxiety, and noted Unity's existing long-term relationship with Google.

Assessment: We find the substance of management's argument more right than wrong. The hard part of a live-service game is not the asset generation, which has always happened largely outside the engine, but the deterministic systems and operations layer that world models do not address, and a larger universe of AI-generated content genuinely does increase the value of a discovery engine. The fatal problem for the rating is that the argument is unfalsifiable on any near-term horizon: world models are early, their trajectory is unknown, and "complementary, not duplicative" is a thesis that cannot be confirmed or refuted by a 2026 print. The market is not pricing management's base case; it is pricing the existence of a fat, unquantifiable tail in the Create distribution. That tail is real, it is new since our upgrade, and we cannot size it, which is the central reason the rating moves to Hold. We do not endorse the existential-threat framing, but we cannot dismiss the uncertainty premium it justifies.

2. The Flat Q1 Grow Guide: Mechanics Versus Optics

The flat sequential Grow guide for Q1 is the fundamental hook on which the de-rating hangs, and management's explanation was entirely mechanical: holiday-rich Q4 seasonality rolling off, two fewer calendar days in Q1, and the continued IronSource decline, with Vector itself still guided to +10% sequentially and Grow expected to return to sequential growth in Q2. The CEO emphasized that the IronSource attention is investor-driven rather than management-driven, calling it a legacy business he highlighted only because "investors are overly focused" on it.

"In Grow, we are forecasting revenue to be flat on a sequential basis due primarily to seasonality as we come off the holiday-rich fourth quarter and with 2 fewer calendar days in Q1. Despite these dynamics, we expect Vector to grow 10% sequentially in the first quarter, and we expect Grow to return to sequential growth in the second quarter." — Jarrod Yahes, CFO

Assessment: The mechanics are credible and the guide is not a Vector-deceleration signal. A flat total is the arithmetic of a +10% Vector line, a shrinking IronSource line, and a tougher seasonal base, and excluding IronSource the segment is still growing. But credibility of the mechanics is not the same as a benign read for the equity. For two quarters the stock ran on a beat-the-conservative-guide pattern; a flat guide removes the visible cushion and converts the pattern's absence into a negative even though the underlying engine is unchanged. The honest read is that Q1 is a mechanically soft quarter the company chose to disclose plainly, landing on the worst possible morning for plain disclosure.

3. No Full-Year 2026 Guidance: The Anchor the Street Wanted and Did Not Get

For a stock being asked to underwrite a multiyear acceleration into a newly uncertain competitive backdrop, the absence of any full-year 2026 revenue or margin framework is conspicuous. Management offered only directional color, that adjusted EBITDA margins should expand 300 basis points year-over-year in Q1 and improve through the year, driving solid full-year margin expansion, and that the non-strategic runoff is roughly $40M for 2026. There was no full-year revenue range, no Vector dollar target beyond the qualitative ">$1B annualized run-rate by year-end," and no segment-level annual framework.

"We expect adjusted EBITDA margins to expand 300 basis points year-over-year in the first quarter… adjusted EBITDA margins should improve throughout the year and drive solid overall margin expansion for Unity in 2026. I would note that we expect healthy margin expansion despite a heavy investment in our product road map across Vector and a range of strategic AI initiatives." — Jarrod Yahes, CFO

Assessment: Continuing to guide one quarter at a time is consistent with Unity's practice all year, and on a clean tape it would be a footnote. On this tape it is a cost. When the central debate has become a multiyear question, whether world models reshape the franchise, the refusal to put any annual anchor on the table leaves the out-year entirely to investor imagination, and imagination under a fresh disruption narrative skews bearish. The 300-basis-point Q1 margin-expansion commitment is concrete and valuable, but margin guidance without revenue guidance answers the question the bears are not asking. We would weight a credible full-year revenue framework as the single most valuable disclosure management could provide next quarter.

4. Runtime Data Goes Live: The 2026 Catalyst Gets a Hard Window

The runtime-data opportunity, feeding differentiated behavioral signal from Made-with-Unity games into Vector, advanced from "infrastructure in place" at Q3 to a concrete launch window this quarter. Management will scale testing of runtime engine data over Q1 with the expectation that it goes live in Vector during Q2, enabled by >90% opt-in on the Developer Data Framework and a new self-service feature that lets games on older Unity versions participate. Consistent with prior quarters, the CEO declined to promise a "lightning strike," framing the contribution as significant, compounding model improvements over time rather than a single event.

"Over the course of Q1, we'll scale our testing of runtime engine data with the expectation that it will be live in Vector during Q2… we don't anticipate that the inclusion of runtime data will produce a lightning strike moment, but rather, it's our conviction that the addition of highly differentiated behavioral data will result in significant, compounding model improvements over time." — Matthew Bromberg, CEO

Assessment: This is the cleanest piece of good news on the call, and it is genuinely thesis-supporting. The catalyst we held at "on track, de-risking" a quarter ago now has a Q2 2026 go-live and a behavioral-data moat that is exactly the kind of proprietary signal the Genie narrative cannot replicate, world models do not generate the runtime engagement data that makes Vector's targeting differentiated. The honest decoupling of Vector's current trajectory from runtime data remains the right framing: the catalyst is upside to a thesis that already works. The irony is that the single most under-discussed item on the call, a proprietary-data moat going live in two quarters, is the best rebuttal to the narrative that dominated it.

5. Commerce / Unity IAP: Early Access Next Week, GA in Q2

The cross-platform in-app-payments product introduced last quarter advances to early access next week and general availability in Q2. Management cited three benefits customers are responding to: faster ability to take advantage of the changing storefront regulatory environment and control their own payment layer, the potential for purchase behavior to enhance Vector models over time, and native integration that makes adoption frictionless. Consistent with Q3, the economics were framed as a very high margin but very modest take, with the strategic point being value and data rather than a near-term revenue line.

"We participate in the economics of the commerce transactions at an extremely high margin but very modest… our goal here is not to make massive dollars on these transactions. It's really to deliver value to customers… and also fundamentally enhance the value of Vector because optimization around engagement and the experiences which lead downstream ultimately to transactions and revenue growth are a really important part… of forming a complete picture of the video game consumer." — Matthew Bromberg, CEO

Assessment: Commerce remains genuine optionality with a clear secular tailwind (app-store disintermediation) and a structural advantage (engine-native integration), and the explicit link to enriching Vector's behavioral picture is the strategically interesting part. It is still not a revenue thesis and management is right not to pitch it as one. We treat it as a free option on the commerce layer and on incremental Vector signal, not a modeled line. The acceleration to GA in Q2 is a modest positive on execution.

6. Create's 2026 Transformation: Browser-Based Authoring and AI Game Generation

Management framed 2026 as a year of fundamental Create transformation along two axes that expand the addressable market. First, collaboration: Unity authoring workflows become largely accessible by web browser with one-click shareable URLs, extending Unity from its current software-developer-only user base to the artists, designers, and product managers around them. Second, AI authoring: at the Game Developers Conference in March, Unity will unveil a beta of an upgraded Unity AI that prompts full casual games into existence with natural language, native to the platform and leveraging frontier models plus Unity's project context.

"At the Game Developer Conference in March, we'll be unveiling the beta of the new upgraded Unity AI, which will enable developers to prompt full casual games into existence with natural language only, native to our platform… This assistant will be powered by our unique understanding of the project context and our runtime while leveraging the best frontier models that exist." — Matthew Bromberg, CEO

Assessment: This is the offensive counter to the Genie narrative, and it is the right strategic instinct: rather than defend the engine against AI creation, embed AI creation inside the engine and use it to expand the user base from software developers to "the many." If executed, browser-based authoring plus native AI generation materially widen Create's TAM and convert the disruption narrative into a growth narrative. The risk is execution and timing, these are betas and roadmap items, not shipping revenue, and the GDC reveal in March is the near-term proof point the Street will watch. We weight it as credible strategic direction with the burden of proof still ahead.

7. Margin Mix and the IronSource Wind-Down as a Profitability Tailwind

Management was explicit that the IronSource decline is not just a revenue-mix story but a margin story: displacing commoditized, lower-margin ad-network revenue with higher-margin Vector platform revenue, plus the ability to concentrate resources currently spread across multiple ad networks into the core engines, yielding greater operating leverage and gross margin. The CFO tied this directly to the 2026 margin-expansion guide, the 300-basis-point Q1 year-over-year improvement and the through-the-year trajectory, and noted Unity reaches "the core growth engines of our business by the end of this year" as IronSource fades.

"We are spreading resources across multiple networks. As our business evolves and changes, we'll be able to ultimately concentrate those resources, leading to greater operating leverage and ultimately greater gross margins in our business… really, we're getting to the core growth engines of our business by the end of this year." — Jarrod Yahes, CFO

Assessment: This is a credible and underappreciated structural positive embedded in the IronSource wind-down that the market is treating purely as a revenue headwind. As the lowest-margin revenue runs off and resources concentrate, both gross margin and operating leverage should improve, which is consistent with the 25% EBITDA print and the 300-basis-point Q1 guide. The same wind-down that makes the Q1 Grow guide optically flat is, on a 12-month view, margin-accretive. We carry this as support for the margin-expansion leg of the thesis even as the revenue optics suffer near-term.

8. Balance Sheet: $2B+ Cash, 2026 Converts Refinanced to 2030

The balance sheet improved further. Unity refinanced $690M of its 2026 convertible notes, extending those maturities into 2030, and exited the year with over $2B of cash ($2.06B). The CFO reiterated that with cash on hand and a highly cash-generative business, the company is confident it can pay off future obligations from the balance sheet and operating cash flow, no capital raise required. This continues the active management of the convert wall that began with the $688M 2026-note repurchase at a gain in Q3.

"We successfully refinanced $690 million of our 2026 convertible notes, extending those maturities into 2030. With over $2 billion in cash on hand and a highly cash flow generative business, we are confident in our ability to pay off future obligations using cash on the balance sheet and cash generated from our business." — Jarrod Yahes, CFO

Assessment: The convert overhang we tracked as "contained, actively managed down" at Q3 is now a non-issue on a 12-month view. Pushing the 2026 maturities to 2030, on top of last quarter's repurchase, with $2B-plus of cash and $400M-plus of annual free cash flow, removes the refinancing-and-dilution concern as a swing factor. This is unambiguously positive and one more reason the operational case at the post-print price is cheap. The balance sheet is not where the risk lives this quarter.

9. Governance: Founder Board Departures and CFO Selling

Two governance signals landed in the same window as the narrative and guide. Both founder-affiliated directors stepped down from the board effective February 5, days before the print: Unity co-founder David Helgason and IronSource co-founder Tomer Bar-Zeev. Separately, the press materials disclosed CFO Jarrod Yahes sold 14,914 shares ahead of the earnings call. Management did not address either on the call. (The board is adding former Match Group CEO Bernard Kim effective May 1, a refresh rather than a net reduction in seats.)

Assessment: Individually, none of these is disqualifying. Founder departures from a company several CEO-transitions past its founding era are not unusual, pre-scheduled 10b5-1 sales are routine, and adding an experienced operator to the board is constructive. But the cluster matters: two founders exiting the board the same week a competitive narrative crystallizes and a soft guide is delivered, with a CFO sale on the tape and a full-year outlook withheld, is a pattern that argues for a higher discount rate on a thesis that already leaned on management credibility. We do not read malfeasance into any single item; we read the aggregate as a reason to demand a wider margin of safety, which the rating now reflects.

Guidance & Outlook

The guide is the cleanest contrast with the Q3 setup, and it cuts the opposite way. A quarter ago the Q4 guide raised the bar, a $485M midpoint above the quarter just reported. This quarter the Q1 midpoint of $485M sits roughly flat-to-below the $503.1M just reported and below the ~$491–494M consensus, which the Street had built on the assumption that Unity would beat and raise as it had twice. The composition is a +10% sequential Vector line offset by the IronSource decline and a holiday-rich Q4 base with two fewer calendar days, netting to flat Grow, plus continued double-digit Create. Adjusted EBITDA of $105–110M holds a ~22% margin and, importantly, is guided up 300 basis points year-over-year even on the seasonally softer revenue, which is the operating-leverage thesis surviving the optical revenue dip.

The qualitative framing was confident on Vector and the full-year margin trajectory and notably silent on full-year revenue. Management characterized the flat Grow guide as pure mechanics (seasonality, calendar days, IronSource), reiterated the >$1B Vector exit run-rate, and committed to full-year margin expansion despite heavy product and AI investment, but declined to put any 2026 revenue range on the table. For a stock whose central question has become multiyear, that omission is the costliest part of the outlook.

Implied sequential setup: The $485M revenue midpoint is roughly flat off Q4's $503.1M, with Vector (+10% sequential) doing the work, IronSource subtracting, and Create growing double-digit year-over-year. Given Vector's demonstrated pattern of exceeding internal expectations, the realistic path likely lands at or modestly above the guided total, but the visible beat cushion that powered the stock the last two quarters is gone for Q1 by construction.

Street at: Consensus entering the print sat near $490–494M for Q1; the guide midpoint of $485M is below it, the first below-consensus guide in this recovery and the proximate fundamental catalyst for the sell-off independent of Genie.

Guidance style: Mechanically conservative on Grow (a flat guide that nets a +10% Vector line against legacy runoff), disciplined and explicit on margin (+300bps year-over-year), and conspicuously absent on full-year revenue. The two-quarter pattern of guiding Grow to mid-single digits and printing double was already broken by the move to a flat framing; we model Q1 Grow at or slightly above flat and watch the Q2 return-to-growth commitment as the key near-term tell.

Analyst Q&A Highlights

Where Vector Is in Harvesting Low-Hanging Fruit, and the IronSource Drag

The opening exchange went straight to the durability question the market was asking that morning: how much of Vector's gain is early low-hanging fruit versus a long runway, and how much of a drag IronSource remains through 2026. Management rejected the premise that there is a near-term ceiling, pointing to the +70% January year-over-year growth, the fact that all of it predates any runtime-data contribution, and the framing of IronSource as a legacy business investors over-weight relative to its shrinking materiality.

Q: "Grow grew double digits organically for the first time in 4 years… what the market is wondering this morning is where are we in the process of kind of harvesting the low-hanging fruit for Vector? And how many significant ongoing breakthroughs are there still ahead for Unity ads? And then how much of a drag is IronSource going to be as we move through the rest of 2026?"
— Matthew Cost, Morgan Stanley

A: "We are just thrilled with the continued strong growth of Vector… January, this is a business up more than 70%… all of this growth predates any of the impact, which we believe will be substantial over the long term of including our runtime data in our models. I know that there appears to be some consternation in the market about the long-term ability for us to grow this business. I honestly have a difficult time understanding why… we feel like there is no natural ceiling to what this business can do."
— Matthew Bromberg, CEO

Assessment: The substance is strong, the +70% January figure with no runtime-data contribution yet is a genuine durability argument, but the CEO's "I honestly have a difficult time understanding why" there is consternation is a tell of a management team frustrated by a stock that has stopped responding to the numbers. The answer grades the Q3 commitment on Vector durability as clearly met. It does not, and cannot, resolve the multiyear question the market is actually pricing, which is not whether Vector grows this year but whether the franchise around it is structurally exposed to world models.

Google Genie and the World-Model Threat

The most consequential exchange of the call was also the bluntest question: a single-sentence ask for the CEO's view on Google Genie and what it means going forward. The response was the longest of the call, a structural argument that world models are complementary asset-and-inspiration sources that Unity's engine operationalizes into real games, not substitutes for the deterministic systems layer that defines a game.

Q: "Matt, I think everyone would love to hear your thoughts on Google Genie and what that means going forward. And that was my question."
— Brent Thill, Jefferies

A: "We believe world models are going to be a source of inspiration and assets for creators, but that they are not in any way going to replace game engines. They are complementary, not duplicative… Unity is not an interactive video generator, it's a 3D execution platform designed to build once and then run everywhere efficiently and seamlessly. So that's my feeling about Genie."
— Matthew Bromberg, CEO

Assessment: That the entire investor anxiety could be compressed into a one-line question, and required the longest answer of the call to address, is the structure of the quarter in miniature. The argument is coherent and we lean toward agreeing with it: the systems-and-operations layer is the durable part of the engine and world models do not build it. But the exchange also demonstrates why the rating moves to Hold, management's best case is a logical thesis about a technology whose trajectory is unknown, and the market is rationally pricing the uncertainty band rather than the base case. A persuasive paragraph is not a quantifiable rebuttal, and on a one-quarter horizon none is available.

Create's Long-Term Strategic Positioning Under the AI Narrative

A follow-up pressed the Create-specific version of the Genie concern: investor worry about the long-term strategic positioning of the engine-subscription business, and what management is actually seeing across the customer base versus the narrative. Management answered with the printed numbers, Create up 16% ex-nonstrategic and accelerating, Unity 6 adopted faster than any release ever, and reframed the time spent on Genie as an opportunity to explain the depth of value the software provides rather than a concession of threat.

Q: "There's a lot of investor concern about the long-term strategic positioning of Create. Maybe you just want to address a little bit what you're seeing across the base of customers in Create today relative to the broader narrative that maybe has sort of made its way into the investor conversations."
— Eric Sheridan, Goldman Sachs

A: "We are seeing incredible strength in our Create business… just a few quarters ago both Create and Grow segments were shrinking. And a little bit more than a year down the line, our largest ad business is growing 70% and the Create business is up 16%… The time I spent answering the Genie question, it's really nothing to do with Genie… The reason I spent the time was to try to explain the depth of value that our software provides to makers of interactive entertainment."
— Matthew Bromberg, CEO

Assessment: The strongest rebuttal to the disruption narrative is the one in this answer, the supposedly-threatened segment is accelerating to its fastest growth in two years with record engine adoption. That is a real, printed contradiction of the bear thesis, and it deserves weight. But the CEO's "it's really nothing to do with Genie" is a rhetorical move that does not survive the fact that he had just spent the call's longest answer on it; the market clearly thinks it has everything to do with Genie. We credit the printed strength and discount the framing.

Competitive Pressure from Meta on iOS Inventory

A question probed whether Meta had become meaningfully more competitive on iOS ad inventory in Q1 and whether that had affected Vector, a concern management attributed to a circulating LinkedIn post. The CEO was direct that Unity has always competed with the largest ad platforms, that Meta's iOS competitiveness is not a new dynamic, and that the impact was essentially nil.

Q: "Could you help us understand what you've seen in the market from Meta so far in the first quarter? Have they become meaningfully more competitive on iOS inventory? And has that impacted the growth of Vector at all?"
— Alec Brondolo, Wells Fargo

A: "I know there was a lot of consternation over the course of the quarter, which as far as I can tell, was kicked off by a LinkedIn post… Meta has been competitive on iOS traffic for quite some time. This wasn't a new dynamic. It did not have a meaningful impact on us in any way… I would, in general, caution investors from overreacting to LinkedIn posts."
— Matthew Bromberg, CEO

Assessment: The substance is reassuring, Unity is laser-focused on games rather than e-commerce, and Meta's iOS presence is long-standing rather than a step-change, and we take the "no meaningful impact" at face value given Vector's printed acceleration. The recurring "overreacting to LinkedIn posts" refrain, deployed on both the Meta and Genie questions, is a window into management's read of the quarter: it believes the sell-off is sentiment-driven noise. That may be partly right, but a management team that attributes a 26% drop primarily to social-media overreaction risks under-weighting the legitimate forward concerns (the flat guide, the absent outlook) bundled with the narrative.

Runtime Data Contribution and Timing

A question sought more specificity on the runtime/Developer Data Framework contribution to the ad model and the timing. Management reiterated the >90% opt-in, the new self-service path for older Unity versions, the Q2 go-live in Vector, and the consistent framing that the benefit is compounding rather than a single step-change.

Q: "You talked about the Developer Data Framework and kind of layering that into the model in 2026. Is there any additional help you can provide us in terms of the contribution from that or maybe a little bit more on the timing as we think about what that could mean for the ad model?"
— Andrew Boone, Citizens

A: "We rolled out the Developer Data Framework first in August… We've had opt-in rates in excess of 90%… We also more recently rolled out a streamlined self-service feature that allows customers that are operating games using older versions of Unity to also take advantage of the Developer Data Framework… we are kind of moving into Q2 release of the runtime data into our models. So again, we plan to do that integration in the second quarter."
— Matthew Bromberg, CEO

Assessment: This grades the Q3 commitment to sharpen the runtime-data timing as clearly delivered, the catalyst now has a Q2 2026 go-live, up from the looser "2026 event" framing of last quarter, plus the self-service expansion that widens the addressable data set beyond new Unity 6.2 projects. Management again declined to quantify the contribution, which is appropriately disciplined for an unproven feature, but the firmer timeline and the proprietary-behavioral-signal angle are exactly the moat the world-model narrative cannot touch. We treat the firmer window as a modest positive and the refusal to quantify as prudent rather than evasive.

Aggregate Grow Trajectory as IronSource Fades

A question sought confirmation of the bridge: a >$1B Vector exit run-rate, IronSource falling from 11% to below 6% of total, and that natural attrition allowing reported Grow to reflect Vector's momentum, plus where the non-IronSource, non-Vector Grow products sit. Management confirmed the characterization and added that, excluding IronSource, the Grow segment grew double digits sequentially in Q4, with all other Grow businesses showing sequential growth.

Q: "I think you said that you're going to exit fiscal '26 at $1 billion plus run rate within Vector. You gave us the 11% moving to 6% on IronSource. So maybe that natural attrition will allow Grow to properly reflect the recent Vector momentum. I guess, is that a fair characterization?"
— Dylan Becker, William Blair

A: "The answer to your first question is yes… not only will that lift growth rates, it will increase profitability because taken as a whole, Vector is a more profitable product… outside of the IronSource Ad Network, all other Grow businesses actually showcased sequential growth in the fourth quarter… excluding IronSource, the Grow segment was up double digits sequentially in the fourth quarter."
— Matthew Bromberg, CEO

Assessment: This is the most useful modeling exchange on the call and the antidote to the flat-guide optics. It establishes that the flat Q1 Grow framing is entirely an IronSource-and-seasonality artifact, that ex-IronSource Grow grew double digits sequentially in Q4, and that the legacy drag falls to sub-6% of total in Q1, after which the reported segment increasingly reflects Vector. The implication is that once IronSource is immaterial, reported Grow re-accelerates to something close to Vector's underlying rate. For modeling, this argues the flat Q1 is the trough of the optical drag, not a fundamental ceiling, and the H2 2026 reported numbers should look materially better than Q1's headline.

What They're NOT Saying

1. No full-year 2026 revenue or segment framework. Management guided margin direction (+300bps year-over-year in Q1, expanding through the year) but provided no full-year revenue range, no Vector dollar target beyond the qualitative ">$1B run-rate," and no segment-level annual outlook, at the exact moment the central debate became a multiyear one. Margin guidance without revenue guidance answers a question the bears are not asking.
2. No quantification of the Genie/world-model risk, in either direction. The CEO argued forcefully that world models are complementary, but offered no framework for what share of Create's TAM is potentially exposed, no scenario analysis, and no metric the Street could use to size or track the risk. The rebuttal was entirely qualitative, which leaves investors to price the tail by imagination.
3. The board departures and CFO sale went unaddressed on the call. Two founder-affiliated directors stepped down effective February 5 and the CFO sold 14,914 shares ahead of the print, yet neither was mentioned in prepared remarks or volunteered in Q&A. For governance events this proximate to a guide-down, the silence is conspicuous.
4. Vector's growth rate is still not disclosed in dollars. Management quantified Vector's % of Grow (56%), its January year-over-year (+72%), and its cumulative since-launch growth (+53%), but the withdrawn dollar-level Unity Ad Network disclosure remains withdrawn, so the Street still cannot independently track the Vector-versus-legacy split on which the entire thesis turns.
5. No dollarization of the AI-authoring or browser-collaboration TAM expansion. The 2026 Create transformation (browser authoring, native AI game generation, the GDC beta) was framed as materially expanding the addressable market, but with zero revenue contribution sized, near-term or modeled. These are real options pitched as growth, but management gave no way to value any of them.
6. The world-model "relationship with Google" was mentioned but not characterized. The CEO referenced a "long-term relationship with Google" and "developing relationships across the space" in the context of Genie, but did not say whether that implies any commercial arrangement, data partnership, or distribution path, leaving an intriguing aside undefined.
7. Commerce economics still unquantified. Unity IAP moves to GA in Q2 with a "very modest" take rate described qualitatively, but no take-rate figure, no addressable volume target, and no contribution timeline were provided, consistent with last quarter but now a quarter closer to launch.

Market Reaction

• Pre-print setup (Feb 10, 2026 close): $29.05, entering the print already down 34.2% year-to-date (from $44.17 at 2024 year-end) and down 38.6% over the trailing 30 days (from $47.29 on January 12), even as it remained up 49.0% over the trailing twelve months (from $19.50). The 52-week closing range was $16.75–$49.47. The critical context is that the Genie-driven re-rating had already done most of the damage before earnings: the stock entered the print roughly 41% below its January peak, so the multiple had been taken apart on narrative alone, ahead of a single reported number.
• Reaction session (Feb 11, 2026, reported before the open): The stock gapped down to open at $21.41 (−26.3%), traded an intraday range of $18.80 to $22.10 (−35.3% to −23.9%), and closed at $21.41, down 26.3% (−$7.64) on the session. The beat-the-high-end print did not arrest the decline; if anything, the flat guide and absent outlook accelerated it.
• Volume: 104.9M shares traded versus a 14.0M 30-day average, a 7.5x surge and a record level of activity, confirming a violent, conviction-driven repricing rather than thin-tape drift.
• Benchmark context: The S&P 500 closed flat (0.0%) on the reaction day and was up 1.4% year-to-date, underscoring that a 26.3% single-day decline was entirely idiosyncratic to Unity, not a market event.

The reaction is the inverse of Q3 and the entire point of the quarter. Three months ago a beat plus a raise plus breadth produced an +18.1% close that held; this quarter a larger beat plus better breadth produced a −26.3% close on record volume. The difference is not the print, which was better, but the forward distribution. By the Feb 10 close the Genie narrative had already compressed the multiple from roughly 9.6x forward sales at the Q3 close to roughly 4x at the pre-print level, repricing the stock for a fat disruption tail before earnings. The print then handed the de-rating a fundamental hook, a flat Q1 guide below consensus and no full-year framework, which converted "narrative overhang" into "narrative plus a real second-derivative slowdown," and the governance signals supplied a credibility discount on top. The 7.5x volume confirms genuine, heavy repositioning rather than a sentiment air-pocket. At the $21.41 close the operational business, growing double digits with a 25% margin and $400M-plus of free cash flow, is demonstrably cheap; the reason the stock is there is that the market is no longer willing to underwrite the out-years at the prior multiple until the world-model question is more legible. That is the precise risk the valuation pillar carried into the print, and it is now the center of the rating.

Street Perspective

Debate: Do World Models Disrupt the Create Franchise or Feed It?

Bull view: The bull case on the Street holds that world models are complementary tooling, not substitutes, that the durable value of a game engine is the deterministic systems-and-operations layer (physics, networking, monetization, live ops) that generative video does not build, and that a larger universe of AI-generated content actually increases the value of Unity's discovery engine. The printed evidence, Create accelerating to +16% with record Unity 6 adoption in the same quarter the narrative peaked, is offered as proof the disruption is theoretical rather than observed.

Bear view: The bear camp contends that world models represent an early but rapidly improving technology that could, over a multiyear horizon, collapse large parts of the game-creation stack into a prompt, that incumbents always rationalize disruptive technology as "complementary" until it is not, and that the right response to an unknowable but potentially existential tail is a structurally lower multiple regardless of near-term numbers. On this read, today's printed strength is irrelevant to a 3-to-5-year displacement risk.

Our take: We lean toward the bull substance and the bear's pricing discipline simultaneously, which is exactly what a Hold expresses. Management's structural argument is more right than wrong, the hard, durable part of an engine is the systems layer, and world models do not build it, and the accelerating Create print is a real contradiction of the live-disruption thesis. But the risk is genuinely unquantifiable on any near-term horizon, the technology is early and its trajectory unknown, so a wider uncertainty premium is rational even if the base case is benign. We do not endorse the existential framing, but we cannot size the tail, and an unsizable tail on the franchise's long-duration segment is incompatible with the conviction an Outperform requires. The single most important thing that would resolve this debate is time and product evidence, neither available this quarter.

Debate: Is the Flat Q1 Guide a Vector Problem or an IronSource Artifact?

Bull view: The bull view argues the flat Q1 Grow guide is purely mechanical, a +10% sequential Vector line offset by a still-shrinking IronSource and a holiday-rich Q4 base with two fewer calendar days, and that management explicitly confirmed ex-IronSource Grow grew double digits sequentially in Q4. With IronSource falling below 6% of total in Q1 and Grow guided back to sequential growth in Q2, the optical trough is Q1 and the reported numbers re-accelerate toward Vector's underlying rate thereafter.

Bear view: The bear camp counters that, whatever the mechanics, a company that beat-and-raised twice and then guided flat-to-below-consensus has broken the pattern that supported the stock, that "seasonality and calendar days" is the explanation every decelerating company offers, and that the refusal to provide a full-year framework means investors cannot verify the Q2 re-acceleration claim. The absence of a beat cushion, on this read, is itself the signal.

Our take: The bull has the better of the mechanics, and the Q&A confirms it: ex-IronSource Grow grew double digits sequentially in Q4, and the legacy drag is nearly spent. The flat Q1 is far more plausibly the trough of an optical headwind than a fundamental ceiling, and the H2 2026 reported Grow should look materially better as IronSource hits immateriality. But the bear's point about the lost cushion is not wrong as a near-term stock dynamic, and the missing full-year framework means the Q2 re-acceleration is a management assertion rather than a verifiable guide. We side with the bull on the fundamentals and the bear on the near-term setup, and would upgrade our conviction on a clean Q2 return-to-growth print.

Debate: At ~4x Sales, Is the Operational Beat a Gift or a Value Trap?

Bull view: The bull view argues the sell-off has overshot: a business growing revenue 10%, expanding margins to 25%, generating $400M-plus of free cash flow at 99% conversion, with a de-risked balance sheet and a Q2 runtime-data catalyst, trading at roughly 4x trailing sales is priced as if it were shrinking, which it demonstrably is not. The Genie fear, on this read, is a sentiment overreaction (the CEO's "LinkedIn post" framing) that a few clean prints will reverse, making the post-crash level an asymmetric entry.

Bear view: The bear camp notes that a multiple does not compress 60% in two weeks on sentiment alone, that the market is efficiently pricing a real (if unquantifiable) long-duration disruption risk plus a genuine near-term guide-down plus a governance discount, and that "cheap on trailing numbers" is exactly how value traps present when the forward distribution has structurally widened. A name with an existential question mark over its long-term franchise, the bears argue, deserves no multiple-mean-reversion until that question is answered.

Our take: This is the crux of the rating, and it is why the call is Hold rather than either Outperform or Underperform. The operational case at ~4x sales is genuinely cheap and the balance sheet removes the solvency tail, which is why we are not downgrading to Underperform, the business is better than the price implies and the cash generation is real. But we cannot underwrite multiple expansion while the Genie tail is unsizable and the near-term guide is softening, which is why we are not maintaining Outperform, a re-rating requires the market to regain conviction in the out-years, and nothing this quarter provides that. A stock this cheap with this much unresolved narrative uncertainty is the textbook setup for a market performer: real downside protection from the valuation and cash, real upside cap from the overhang, netting to roughly in-line over the next twelve months until the world-model question gains legibility.

Model Update & Valuation

The operational lines in our framework move up this quarter; the valuation and risk-weighting lines move materially. The table frames the changes against where we carried the model after the Q3 upgrade.

Valuation framework: At the reaction-day close of $21.41, Unity trades at roughly 4x trailing sales, a level last seen when the franchise was reporting shrinking revenue, against a business now growing 10% with a 25% margin and $400M-plus of free cash flow. On the operational numbers alone, that is cheap, and a quarter ago, with the valuation pillar the only live risk at a 9.6x multiple, this de-rating would have been the entry we were waiting for. What changed is not the price but the distribution behind it. The Genie world-model narrative introduced an unquantifiable tail risk to the Create franchise that did not exist at our upgrade, the flat Q1 guide and withheld full-year framework removed the visible forward cushion and the anchor the Street needed, and the founder board departures plus CFO selling added a credibility discount. We can no longer underwrite the multiple expansion an Outperform requires, because that expansion is contingent on the market regaining conviction in out-years that are now subject to a narrative no print this year can resolve. Equally, we cannot justify an Underperform, because the operational case is demonstrably cheap, the balance sheet removes the solvency tail, and the disruption thesis is, in our judgment, more wrong than right on the merits even if unquantifiable on the timeline. Fair value on a 12-month horizon is roughly the current area, with the valuation and cash providing downside support and the overhang capping upside, until the world-model question gains legibility.

Risk/reward: The up-case is that the Genie fear proves to be the sentiment overreaction management says it is, the Q2 runtime-data go-live and the GDC AI-authoring reveal reframe AI as a Unity tailwind rather than a threat, Grow re-accelerates as IronSource hits immateriality, and the stock re-rates off a washed-out ~4x base. The down-case is that world models improve faster than the bull case assumes and the disruption narrative hardens from sentiment into a structural de-rating, or that the flat Q1 is the first of several softening quarters and the absent full-year framework was hiding a slowdown. The difference from a quarter ago is that the one live risk we named, valuation, broke through a channel we did not anticipate (a narrative re-rating rather than a fundamental miss), and a new, unsizable risk (world-model disruption) joined it. That combination, cheap operations against an unquantifiable tail and a softening near-term guide, is the definition of a balanced risk/reward, and the basis for the downgrade to Hold.

Thesis Scorecard Post-Earnings

This scorecard grades the same pillars carried in the standing thesis against what Q4 revealed, and grades the eight specific commitments management made on the Q3 call. The pillar status tags move where the quarter justifies it, and a new bear pillar is added for the world-model disruption risk that did not exist at the Q3 upgrade.

Overall: The thesis is operationally strengthened and risk-profile-weakened at the same time, an unusual and uncomfortable combination. All four bull pillars are confirmed (three strengthening, runtime data de-risked to a Q2 go-live), and two of the three original bear pillars (single-product dependency, balance sheet) are further challenged toward resolution. But the one original bear pillar that was live, valuation, reset hard through a narrative-driven re-rating rather than a fundamental miss, and two new bear pillars emerged that did not exist at the Q3 upgrade: an unquantifiable world-model disruption tail to Create, and a governance/forward-visibility discount. The net of a better business and a structurally wider risk band is a thesis that no longer supports the conviction an Outperform requires.

Action: Downgrade to Outperform's successor rating, Hold. We are not selling the operational story, which delivered nearly everything we asked and prints cheap at ~4x sales with a de-risked balance sheet and a Q2 catalyst. We are declining to underwrite multiple expansion while an unsizable disruption tail and a softening near-term guide sit on the name, and while the governance signals argue for a wider margin of safety. The path back to Outperform runs through three observable gates: a clean Q2 Grow return-to-growth print, the runtime-data go-live and GDC AI-authoring reveal reframing AI as a tailwind, and either a full-year framework or simply the passage of enough time to make the world-model question legible. Until then, cheap operations against an unquantifiable tail is a Hold.