Vector Inflects, the Stock Front-Runs It: Pre-Market +12% Reverses to -6% Close

Results vs. Consensus

Year-over-Year Comparison

Quarter-over-Quarter Comparison

Revenue Assessment

The $14M revenue beat is the cleanest top-line beat Unity has posted in any quarter since the ironSource merger closed. The composition is what matters: of the ~$15M upside vs. guide, the bulk came from the Unity Ad Network running faster than the company had modeled at the start of Vector's rollout. Management explicitly flagged that the Unity Ad Network is now ≈50% of total Grow revenue and growing 15% sequentially, which means the segment-level math is increasingly carried by a single asset that didn't fully exist in this form three quarters ago. That concentration cuts both ways: it's the basis for the bull case (we are early in a real product cycle on a high-gross-margin business) and the basis for the bear case (the rest of the Grow portfolio, ≈50% of Grow, is still declining in absolute terms or stabilizing at best — the "stabilization" framing is management's, not ours).

The Q3 guide of $440–450M (midpoint $445M) implies essentially flat-to-modestly-up sequential revenue. The composition under the hood is more interesting: Grow up mid-single-digits sequentially, Create down slightly sequentially as the $12M Q2 term license doesn't repeat and a "large customer win" lapping creates a tough comp. If you back out the Create dynamic, the underlying strategic-Create line grows high-single-digits YoY in Q3. We think the guide is a deliberate sandbag on the consolidated number — Bromberg said management has "the benefit of having been through the month of August" and described being "consistently surprised" — but the optics matter, and the midpoint sitting below ~$447M consensus was sufficient to flip the after-print narrative.

Margin Assessment

Adjusted EBITDA margins of 21% on $441M, up from ~18% on $449M a year ago, is a real operating leverage signal: margin expanded on lower revenue, which is rare and only happens when the cost base has been restructured. The 83% adjusted gross margin is the structural ceiling that makes this a high-quality business model when revenue is growing — every incremental dollar of Grow revenue throws off ~80 cents of contribution. The CFO was explicit that the company is now investing into growth while sustaining margin expansion ("we're seeing substantial and sharp improvement in margins as we go") and we believe him given the disciplined Q1 and Q2 headcount picture. The Q3 EBITDA guide of $90–95M on $440–450M revenue implies a midpoint margin of 20.8%, suggesting the company is choosing to reinvest some of the operating leverage into Vector road-map acceleration and AI-in-Create investments rather than flow it all to the bottom line. That is the right capital-allocation choice for a company in product-cycle inflection, even if it dampens near-term EPS optics.

EPS & Cash Flow Assessment

Adjusted EPS of $0.18 vs. $0.13 consensus is a clean operating beat — no tax-rate weirdness, no below-the-line magic, no share-count distortion. GAAP EPS of $(0.26) reflects $130M of SBC and $90M+ of intangible amortization from the ironSource deal; both of those drags persist on a slow burn-down schedule. The story to focus on is FCF: $127M is the highest quarterly FCF since the IPO, and FCF/revenue at 29% is in the top decile of software peers. Some of that is publisher-payment-timing flattering, but the trajectory through H1 ($72M Q1 + $127M Q2 = $199M H1 FCF, run-rated $400M for the year on $1.75B of revenue is ~23% FCF margin) is structurally sound. Cash balance of $1.7B against $2.2B of convertible debt leaves net debt of ~$500M, which on $400M FCF is a 1.25x net-debt/FCF — a non-issue. The capital structure is no longer a thesis variable.

Segment Performance

Create Solutions — $154M, +2% YoY, +2.7% QoQ

Create is doing two things simultaneously and the headline number obscures both. First, the underlying strategic subscription business is growing double-digits for the second straight quarter, and after normalizing for the $21M of nonstrategic Create revenue that lapped out of the Q2'24 base, the strategic-Create line grew 16% YoY. That is a structurally faster rate than Create has shown since 2022. Second, the $12M term license contract embedded in this quarter is a one-time recognition that doesn't recur — strip it and reported Create was $142M, which is -6% YoY on a headline basis. Management was deliberate in flagging this both because they want the strategic-growth narrative to land cleanly and because they don't want investors to baseline $154M as the new run-rate when Q3 will come in below it on the lapping effect.

Unity 6 adoption is the leading indicator. 6.6M downloads is +50% QoQ, which suggests the engine generation that the company has been positioning as "the most stable and performant version of Unity we've ever shipped" is being adopted faster than the prior major-version transitions. The follow-through implication is that subscription seat conversions accelerate over the next 2–3 quarters as developers move active projects onto Unity 6 and the AI investments in the editor (announced for Unity 6.2 this summer) become net-new value drivers rather than just retention-defense.

"Our rededication to quality, stability and improving the developer experience is catalyzing strong financial results in Create, including another quarter of double-digit subscription growth in Q2. We're also seeing momentum around the continued adoption of Unity 6, the most stable and performant version of Unity we've ever shipped, which has now registered more than 6.6 million downloads, up 50% from last quarter." — Matt Bromberg, CEO

Assessment: Create is the steady compounder under the Vector noise. The 16% strategic YoY growth is the real number — the headline 2% is a transition-quarter optical artifact. The Q3 guide for slight sequential decline in Create is mathematical (no $12M term license, large customer win lapping); the underlying business is still accelerating. Industry/non-gaming as the fastest-growing subscription line for 10 consecutive quarters is the unmodeled call option — it's small enough to ignore today, but the BMW + Mercedes-Benz + Specto Medical disclosures show it's broadening, not narrowing.

Grow Solutions — $287M total, -4% YoY, +0.7% QoQ (but the QoQ figure hides what matters)

Grow Solutions reported $287M, which on the surface looks like another quarter of flat-to-down ads revenue at a company two-and-a-half years into a turnaround. But the surface number is now structurally misleading because Unity Ad Network — driven by Vector and now ≈50% of total Grow — grew 15% sequentially while the other ad products declined. That mix shift is the entire story for the segment going forward: as the Unity Ad Network's share of Grow grows from 50% toward 70–80% over 2026–2027, the consolidated Grow growth rate converges toward the Unity Ad Network growth rate. The math is straightforward and management is leaning into it. Whether the rest of Grow ever returns to growth is, in management's words, "a road map for clear returns back to growth for each of those products" — translation: each product needs to be re-platformed individually, and that takes 4–8 quarters per product.

Sequential growth in the Unity Ad Network of 15% in Q2 followed by "double-digit" expected sequential growth in Q3 implies a >25% Unity-Ad-Network sequential growth across two quarters, against a structural backdrop where Vector has only been fully rolled out since May. The CEO's language — "we are just scratching the surface" — is consistent with the data we have: a new neural-network-based platform that can ingest more data, more features, more signals, and is learning quickly. The data-framework runtime-data lever ("we anticipate seeing the impact of our work in this area beginning in 2026") is the second-stage rocket. We treat that as bullish optionality, not modelable revenue, for now.

"Vector-led performance inside the Unity Ad Network sparked 15% sequential revenue growth in the second quarter, growth which we are seeing continue into the third quarter as well... we are seeing 10% or more sequential growth on top of the 15% sequential growth we saw the preceding quarter, or at least 25% growth inside of a couple of quarters." — Matt Bromberg, CEO & Jarrod Yahes, CFO

The cannibalization debate — whether Unity Ad Network growth is incremental dollars or simply pulling spend from Unity's other ad products — is the central bear-case framework. Management said cannibalization is "less than 10%" of Unity Ad Network growth. We can't independently verify that number, but the construct management offered (advertisers spend up to their return requirements, and Vector is delivering better returns, so spend expands rather than rotates) is consistent with how performance-advertising spend behaves empirically. We assign moderate but not high confidence to the <10% figure.

Assessment: The Grow story is "is the Unity Ad Network's growth incremental, and how long can it compound?" The headline -4% YoY is the wrong frame. The right frame is: a $144M-quarter product growing 15% sequentially with another 10%+ sequential growth guided into Q3, on a high-gross-margin business model, with an unmodeled runtime-data tailwind in 2026. If we believe the <10% cannibalization figure (we tentatively do), Grow exits 2025 at a high-teens YoY growth rate and re-rates the franchise multiple. The watch item: any quarter where Unity Ad Network sequential growth decelerates faster than expected without a corresponding pick-up in non-Vector ad products is a thesis-break candidate.

Key Topics & Management Commentary

Overall Management Tone: Management was high-conviction throughout, framing Q2 as "an inflection point" without hedging the framing — language that a year ago would have read as overreach but in this print is anchored by the 15% sequential Unity Ad Network number. The CFO was tightly quantitative on the Grow mix (≈50% Unity Ad Network, ≈50% other), the EBITDA-margin trajectory (83% adj gross margin, structural operating leverage on incremental revenue), and the Q3 guide composition (Grow +5% sequential, Create slightly down on $12M term-license lapping). Where the call notably leaned forward versus prior quarters: AI in Create — management committed to "substantially ramping our ambitions for the role AI will play in the core Unity content creation experience" with a tone that read as a deliberate signaling of an upcoming product narrative for 2026. The only soft spot was the response to questions about the pace of extending Vector into non-Unity-Ad-Network products, where the answer was "gradual and constant" rather than a calendar — appropriate honesty, but it leaves the other ~50% of Grow without a hard catalyst date.

1. Vector as a Generational Upgrade — Not a Feature Release

The CEO's framing of Unity Vector throughout the call was that it represents a step-change in the technical architecture of the company's performance-advertising stack, not a tactical feature improvement. The 15% sequential growth in Unity Ad Network is the first-quarter readout of that architecture; the second-quarter readout is the 10%+ sequential growth that management embedded in the Q3 guide.

"Vector was a really important inflection point for our business. And we implemented what was effectively a generational upgrade of our capabilities and we moved them to a brand-new neural network-based platform. It's more powerful. It's more versatile. It's more scalable than our old systems, which means that it can handle more data, can handle more complex types of data, and it can respond in real time to changes in data." — Matt Bromberg, CEO

The architectural framing matters because it sets up the recurring-improvement narrative. If Vector is a one-time level-set, the 15% sequential growth is a one-time benefit and the run rate normalizes. If Vector is a platform that continues to learn — more data, more features, more signals — the sequential improvements continue over multiple quarters at decelerating but durable rates. Management's commitment to the second framing is unambiguous; the question is whether the data supports it across Q3 and Q4.

Assessment: The "generational upgrade" frame is what justifies a higher multiple if Q3 and Q4 deliver. It also means a deceleration in any quarter is a thesis test, not a guidance miss. We are calibrating against Q3 sequential Unity Ad Network growth needing to be at least 10% (per management's commitment); any number below 7% would be a meaningful negative signal.

2. The Developer Data Framework — The 2026 Runtime-Data Lever

The introduction of the Developer Data Framework with Unity 6.2 was framed as the foundational infrastructure for Unity's biggest structural competitive advantage — the runtime data generated by the billions of consumers using Made-with-Unity applications. Management was careful to position it as a control surface for developers (privacy dashboards, opt-in toggles) rather than as a data-acquisition mechanism, but the strategic implication was made explicit: this is the plumbing that lets Vector's AI access "new and highly differentiated behavioral data" in 2026.

"The foundation for this next generation of Unity is being launched this summer with the release of Unity 6.2, which includes the introduction of the developer data framework... We anticipate seeing the impact of our work in this area beginning in 2026 and extending well into the future." — Matt Bromberg, CEO

The reason this matters strategically: every other large performance-ad platform (Meta, Google, ByteDance, AppLovin) competes for the same auction inventory using similar bidding machinery. Unity's structural angle, if the data framework rolls out as described, is access to consumer behavioral signal that flows through the engine itself — what users do inside the games, not just what they did in the ad funnel. That is a different feature set than competitors can replicate without an equivalent install base of in-game telemetry. The timeline ("begins in 2026") is conservative by intent.

Assessment: This is the most important unmodeled lever in the story. If the data framework delivers in 2026 as described, it represents a step-change in Vector's quality, not a marginal improvement. We do not model anything for it today, but it is the basis for our willingness to pay a higher multiple if Q3 and Q4 confirm Vector's near-term trajectory.

3. The Cannibalization Question — <10% per Management

The single most-asked question on the call, across multiple analysts in different framings, was whether Unity Ad Network growth is incremental dollars or simply pulling spend from Unity's other ad products. Management's answer was unambiguous: cannibalization is "less than 10%" of the Unity Ad Network growth, and the conceptual framework was that performance advertisers don't allocate budget in the traditional sense — they spend up to the limit of their return requirements, and Vector is expanding those limits.

"We are quite clear that the cannibalization that there is really no meaningful cannibalization impact at all. It's important to understand that our ad products operate in a big, broad competitive marketplace with some of the largest, most sophisticated companies competing every day. It is not Unity against Unity. So spend flows to where the return is... we estimate that the cannibalization of our other ad network to Unity is less than 10%." — Matt Bromberg, CEO

The framework is consistent with how performance-advertising budgets behave at scale. Where it could be wrong: if a meaningful portion of Unity Ad Network's growth comes from advertisers shifting from Unity's older ad products to the new platform because the older products have degraded relatively (resource reallocation during the Vector launch was explicitly acknowledged), then "cannibalization" is technically intra-firm even if it's incrementally net-positive to Unity. The Q3 guide leans on the non-Vector ad products "stabilizing," which would resolve that ambiguity if it happens.

Assessment: We tentatively credit the <10% figure but assign it moderate, not high, confidence. The cleanest test is Q3 — if non-Vector Grow stabilizes (per management's guide) while Unity Ad Network continues to grow double-digit sequentially, the cannibalization concern is empirically dismissed. If non-Vector continues to decline materially while Unity Ad Network grows, the cannibalization narrative remains a live debate.

4. Major Partnerships — Tencent, Scopely, Nintendo

The CEO disclosed three new multi-year platform partnerships in the prepared remarks: Tencent (one of the largest game developers globally; expanded multi-year deal across CCG/Create titles), Scopely (creator of Monopoly Go!; long-term technical partnership across both Create and Grow), and Nintendo (Unity 6 day-one optimization for Switch 2). These are not seat-license deals — they are platform integrations where Unity becomes the orchestration layer for the partner's broader portfolio.

"With Tencent, our expanded multiyear partnership will keep Unity at the core of some of the most popular multi-platform titles in the world for many years to come... With Scopely, we're embarking on a new multiyear agreement that includes a long-term technical partnership across both Create and Grow... [With Nintendo,] our collaboration ensured that Unity 6 would be fully optimized on day 1 for the release of the Switch 2." — Matt Bromberg, CEO

The strategic framing — "for the first time, I think, maybe in our history, we've been able to begin to package the full value of our platform for customers" — is the right answer to a question that has dogged Unity for years: why isn't a company with this kind of share of game-engine usage able to monetize the platform position more aggressively? The answer historically was that Unity's go-to-market was seat-by-seat. The shift toward platform-level deals with Tencent/Scopely/Nintendo is the structural change that lets Unity capture more of the value chain.

Assessment: Platform-level partnerships are the right strategic shift; the financial impact is back-end loaded and won't show up cleanly in 2025 numbers. Watch for ARPU lift in the strategic-Create base over the next 2–3 quarters as the Tencent/Scopely structures roll into the recognized revenue mix. The Nintendo Switch 2 angle is the most immediate — Switch 2 is the highest-velocity console launch of 2025, and Unity being the day-one engine of choice is a brand-defining moment.

5. AI in Create — The Next Investment Lever

Management committed to "substantially ramping our ambitions for the role AI will play in the core Unity content creation experience" within existing financial parameters. The framing was that Unity's position as the orchestration layer for content creation gives the AI access to context that competitors can't replicate — the platform is "context aware of the project," which lets the AI predict developer needs in a way that a generic creative-AI tool cannot.

"With Vector now successfully underway, you'll see us substantially ramping our ambitions for the role AI will play in the core Unity content creation experience. Within the financial parameters we've already established, we're making significantly increased investments in talent and product in this area and expect to be talking about it much more in the quarters ahead." — Matt Bromberg, CEO

The "within existing financial parameters" qualifier is important — management is not telegraphing a margin reset to fund AI-in-Create. The 21% adj EBITDA margin in Q2, with operating leverage continuing into Q3 (guide implies 20.8% midpoint), is the budget envelope for this investment.

Assessment: AI-in-Create is the under-discussed strategic story. If executed, it deepens Unity's moat in the engine business by raising the cost of switching to a competitor — the AI gets better as more developers use Unity, creating a network effect on the editor side that mirrors what Vector is building on the ad side. We don't model anything for it but treat it as the third pillar (after Vector and Developer Data Framework) of why this could be a different company by 2027.

6. China Strength — $20M Sequential, Broad-Based

The CFO disclosed that China revenue was up "about $20 million sequentially," with growth across both Create (engine adoption + OpenHarmony support) and Grow (advertising spend from Chinese publishers). The Tencent expanded partnership is the marquee, but the broader read is that Unity has structural China advantages competitors lack — multi-platform support across the Chinese mobile OS ecosystem, including OpenHarmony.

"Our business in China was up about $20 million sequentially, which is truly very exciting. Some component of that growth was due to growth in the Create element and other parts of the growth was due to growth in Grow. And so we are seeing broad-based growth in China across both the platform side of our business as well as the advertising side of our business." — Jarrod Yahes, CFO

Assessment: $20M sequential is meaningful — that is ~4.5% of total revenue from one geography. China-spend dependence introduces geopolitical-risk exposure that we monitor, but at the current run rate it remains net-positive to the thesis. The OpenHarmony angle is genuine structural differentiation that justifies a premium relative to Western-platform-only ad-tech peers.

7. Net Revenue Retention Back at 100

Dollar-based net expansion returned to 100 for the first time in several quarters — the metric had been below 100 through 2023 and most of 2024 during the post-IPO digestion period. Management framed the upside potential as predominantly Grow-driven: customer spend increases in performance advertising have higher elasticity than Create subscription price increases, which management intends to keep "modest and stable."

"We are pleased with what we're seeing in terms of net revenue retention, which is the disclosure we've made. We've seen it go up quarter-on-quarter for the last several quarters. That's really a function of the underlying business health being there, both in Create and Grow because that metric is ultimately an output that comes from improvements in ARPU for our existing Create customers as well as improvements in spend for our grow revenue business." — Jarrod Yahes, CFO

Assessment: NRR back at 100 marks the formal end of the ironSource-digestion phase. The next milestone is NRR > 110, which would require Vector-driven Grow spend expansion to accelerate while Create maintains its current trajectory. The Q3/Q4 cadence will tell us whether that is on the near-term horizon or a 2026 story.

8. Operating Leverage on 83% Gross Margins

The CFO returned multiple times to the 83% adjusted gross margin as the structural anchor of the business model, with incremental contribution margins likely higher. The implication: every dollar of incremental Grow revenue throws off ~80 cents of contribution, which can either flow to EBITDA or be reinvested into product. Management explicitly chose to reinvest most of the Q2 leverage into Vector road-map acceleration and AI-in-Create, while still expanding EBITDA margins ~300bps YoY.

"The company is blessed with very high contribution margins. Our adjusted gross margin in the business is 83% with the incremental contribution margins probably higher than that. We expect fully and are beginning to see meaningful operating leverage in our business." — Jarrod Yahes, CFO

Assessment: The 83% gross margin justifies a software multiple, not an ad-tech multiple. The combination of high gross margins, high FCF conversion, and a return-to-growth posture is what unlocks the next multiple re-rating — but it requires growth to be sustained, not just inflected.

Guidance & Outlook

The guide composition is more important than the headline. Grow at "mid-single-digit" sequential growth implies $300M+ Q3 Grow, with the Unity Ad Network sustaining 10%+ sequential while non-Vector ad products stabilize. Create slightly down sequentially is mechanical — no $12M term license, large customer win in Q2 doesn't repeat — but the underlying strategic Create line is up high-single-digit YoY. Adjusted EBITDA at 20.8% midpoint margin reflects continued reinvestment into Vector road-map and AI-in-Create rather than full margin flow-through.

Implied H2 ramp: Q3 guide midpoint $445M + Q4 implied ~$455M (low double-digit YoY exit) = ~$900M H2, or $1,775M FY (≈+1% YoY vs $1,786M FY24 baseline). The FY trajectory is essentially flat, but the exit run-rate is a different rate than the entry.

Street at: Q3 consensus ~$447M (above guide midpoint; in-line with high end). Q4 consensus pre-guide ~$460M; will likely come down post-guide to ~$450–455M.

Guidance style: historically, Unity has guided conservatively under Bromberg — beating both revenue and EBITDA in three of the last four quarters with an average revenue beat of ~3–5% and an average EBITDA beat of ~$10–15M above the high end of guide. The Q3 guide pattern is consistent with that posture; we expect Q3 to print in the $450–460M range with EBITDA around $95–100M.

Analyst Q&A Highlights

Modularity of Vector and Pace of Improvement

The opening question of Q&A pressed on whether Vector's architecture can be ported to Unity's other ad products and how quickly. Management's response was unambiguous on the technical answer (yes, the system is modular by design) but deliberately vague on the timeline. The follow-up asked about pace of improvement specifically — whether Vector's incremental quarterly lifts compound or decelerate — and management framed the answer in terms of model-learning durability, citing the "15% to 20% lift in installs and user value" from Q1 and the continued increase above that range in Q2.

Q: "In terms of just the rate of improvement with Vector — Vector was fully rolled out [in Q2]. What is the pace of improvement that we can expect from Vector over time? It seems like we're not even necessarily going to see the benefits of runtime data until 2026. So I mean how much low-hanging fruit is there left to drive these sequential improvements just in the performance of the model?"
— Matthew Cost, Morgan Stanley

A: "Vector was a really important inflection point for our business. And we implemented what was effectively a generational upgrade of our capabilities and we moved them to a brand-new neural network-based platform... These are capabilities that will continue to grow as the model continues to learn because we can process more features, both dense and sparse, and we can find important signals in what would otherwise be a vast sea of noise for us. So we expect that this investment, which, to your point, is really just very much at the beginning, will continue to provide lifts for us for years. And indeed, even in the immediate term, we are already seeing lifts above the 15% to 20% improvements that we saw in installs and user value last quarter."
— Matt Bromberg, CEO

Assessment: This exchange anchors the bull case. Management is explicitly framing Vector as a multi-year compounding lever, not a one-time level-set. The data point that lifts in Q2 were above the 15–20% Q1 lifts in installs and user value is the most concrete metric on the call. If that trajectory holds for two more quarters, we have a multi-quarter re-rating opportunity. If it decelerates sharply, the "generational upgrade" frame breaks.

$12M Create Term License — One-Time or Pattern?

The CFO was pressed on the nature of the $12M term license recognized in Create — specifically whether it was an anomaly or part of a recurring contract structure that will repeat. The answer was a deliberate threading of the needle: the contract structure is part of broader strategic-partnership deals (Tencent, Scopely, others not named) that include upfront license recognition with majority-SaaS continuation, but management deliberately called out the $12M as a one-quarter benefit because Create will be down sequentially in Q3 without it.

Q: "On the $12 million perpetual deal in Create — can you just describe what happened there? Why that's not recurring? What you saw there that maybe just doesn't [repeat] — is an anomaly? Give us a sense of the color of the background."
— Brent Thill, Jefferies

A: "We're excited about some of the larger partnerships that we signed during the quarter. Matt mentioned several of them, namely Tencent, Scopely. There are several others that we didn't mention by name. Those contracts are fairly far-reaching... there is an upfront recognition of the revenue component of them with the majority of that revenue being a traditional SaaS-based subscription revenue that continues to benefit the business over time. We felt necessary to call out [the $12M] as one component of one of the contracts. It is less than the totality of the contract by far. It's a small component of it just because it did positively benefit the second quarter and wouldn't continue on into the third quarter."
— Jarrod Yahes, CFO

Assessment: The disclosure posture is appropriate — if anything, more conservative than required by GAAP, since the $12M is contractually a real revenue event. The deeper signal is that platform-level partnerships are starting to have non-trivial revenue tails (the $12M is the one-time portion of presumably a $30–50M total contract value). That is a structurally different revenue mix than the seat-by-seat Create business of 2022–2023.

Cannibalization and Non-Vector Stabilization

The dominant theme across multiple analysts on the call: whether Unity Ad Network growth is incremental dollars or rotating from Unity's own other ad products. Management addressed this in three places — the prepared remarks, the opening response on Vector modularity, and most directly in the exchange below. The CEO put a specific number on cannibalization (<10%), and the CFO cross-referenced the Q3 guide assumption of stabilization in non-Vector ad products as the empirical test.

Q: "The debate on the stock this morning is the extent to which the better Unity Ad Network growth in the second quarter was cannibalizing ironSource spend as opposed to incremental to Unity corporate or the business overall. And so maybe any thoughts on that concept — kind of incrementality relative to cannibalization — would be helpful. And perhaps a sense check on your level of confidence that as ironSource or the non-Vector portion of Grow revenue starts to stabilize, we won't see a deceleration or deterioration in the improvement in the rate of growth of the Unity Ad Network."
— Alec Brondolo, Wells Fargo

A: "We are quite clear that the cannibalization that there is really no meaningful cannibalization impact at all... It is not Unity against Unity. So spend flows to where the return is. And we estimate that the cannibalization of our other ad network to Unity is less than 10%... we're seeing ongoing continued significant growth in the Unity Ad Network and stabilization in some of the non-vector ad elements. And so it's really starting to come through for us, and we're really excited about what we're seeing."
— Matt Bromberg, CEO & Jarrod Yahes, CFO

Assessment: The <10% figure is the load-bearing data point for the bull case. We tentatively credit it but want Q3 to provide the empirical confirmation — non-Vector ad products stabilizing sequentially while Unity Ad Network grows 10%+ would resolve the question decisively. If non-Vector continues to decline sharply in Q3, the cannibalization narrative remains a live debate and the multiple compresses.

Mediation Strategy and the AppLovin Comparison

A pointed question on whether Unity needs to win in ad mediation to win in performance advertising — a structural debate that has tracked the AppLovin-vs-Unity competitive frame for years. Management was unambiguous that Unity does not need mediation share to prosper, citing the first-party relationships with billions of end users and game creators as a separate competitive moat. The implicit critique of mediation-led platforms (that they lock customers in through user-acquisition dependence) was sharper than management has been on this topic in prior quarters.

Q: "One of the big questions or debates with investors seems to be over the strategic importance of mediation and collecting signal on ad pricing and inventory. Can you talk about that as part of your Unity Grow offering and whether you see deeper integration of mediation with other growth features like the DSP is unlocking more potential in inventory and sales?"
— Richard Kramer, Arete Research

A: "We believe that we are uniquely positioned in this marketplace, as I've described, by virtue of being providing the operating systems for games globally. So it's not our intention that mediation isn't helpful. It's just that — it's our contention that we don't need to win in mediation in order to prosper and grow the business. We have first-party relationships with billions of customers already... I'm not at all sure that the future of mediation ought to be a system which locks customers into mediation by virtue of their use of user acquisition."
— Matt Bromberg, CEO

Assessment: The strategic positioning here is differentiated and we think defensible — Unity is competing as a platform with proprietary runtime data, not as a mediation layer that needs to win an auction-position fight. The risk is that if mediation winners (AppLovin's MAX, Google's bidding) end up controlling enough of the high-quality inventory, Unity Ad Network's growth ceiling gets capped regardless of Vector's quality. This is the central long-cycle competitive question and management's answer was clear-eyed about it.

Vector Scalability and the Channel-Check Pushback

An analyst surfaced what they called "channel-check" feedback from advertisers pushing back on Vector's scalability — that performance improvements are real but capacity to absorb more spend at those returns is limited. Management's response was both substantive and unusually open about the day-to-day operational reality: yes, scalability is a real constraint for some customers; no, that doesn't change the long-run thesis because the work to expand scalability is the same work as improving model quality.

Q: "Can you talk a bit about the scalability of ad spend on Vector? We've heard in our checks about the strong performance improvement, but we've also heard a bit of pushback on Vector on the scalability side of things. So is there an opportunity to invest in product to improve ease of spend?"
— Chris Kuntarich, UBS

A: "Listen, just to echo the point that Jarrod just made, it's important to remember that we are just a very few short weeks into having rolled out the system. And the answer — the short answer to your question is, yes. While we're really pleased with the results we're seeing, we also see nothing but work ahead of us. I mean we are just scratching the surface here. So yes, many customers are seeing lifts. And yes, those lifts are being seen across genres and across geographies, but there are also some customers that have not yet seen those lifts or where those lifts have hit a point of diminishing returns where they can't get the scale they want. Every advertiser wants infinite scale at their return requirements. And we want to deliver those, but that's work that needs to be done."
— Matt Bromberg, CEO

Assessment: Honest answer that addresses the real concern. The fact that management is willing to acknowledge customers who aren't seeing Vector lifts at scale is a maturity signal — it tells us the company is being calibrated, not hyperbolic, about the rollout. The bear-case extraction here: if scalability constraints are real, the 25%-in-two-quarters sequential growth in Unity Ad Network is structurally capped at some point by the same advertiser-return ceilings management referenced. The bull-case extraction: if model quality continues to improve, the return ceiling moves up, and the spend follows.

Net Revenue Retention and the Path to 110+

A direct question on whether the NRR-at-100 datapoint can move structurally higher and what the components are. The CFO's answer was the cleanest framing of the business-model architecture given on the call: NRR improvement going forward will be driven almost entirely by Grow (advertiser-spend expansion as Vector compounds), with Create contributing only "modest and stable" price/usage growth. This is a deliberate signaling choice.

Q: "Nice to see dollar-based net expansion get back to 100. The changes in the composition of the Create revenue base and the new Vector model — I just wondering how you could characterize DB&E across the 2 businesses. And when we look at the upside potential, are you thinking something in line with what we saw towards the end of '22 in the data sheet? Or is there potentially a higher upside at a steady state?"
— Parker Lane, Stifel

A: "The scale of our Grow business is such that improvements in that metric is really going to come from customer spend increases in Grow. The pace of price improvements in Create is going to be modest and stable. We expect to be able to give our customers predictability and stability. And so you would expect that the improvements in that metric will largely be driven by meaningful improvements in our Grow business and meaningful improvements in customer spend through our platform."
— Jarrod Yahes, CFO

Assessment: The Create posture of "predictability and stability" is a developer-trust signal — Unity is choosing not to optimize for short-term ARPU because the long-term relationship matters more. NRR upside being Grow-driven is the right architecture but means NRR is now structurally indexed to Vector's performance, with no diversification from the Create line. If Vector underperforms in any quarter, NRR moves with it.

Intra-Quarter Vector Trajectory

The final question of Q&A surfaced the pacing of Vector growth through Q2 and into Q3. The CFO's response disclosed the most data ever offered on intra-quarter trajectory: each month of Q2 was better than the last, and that strength continued in July and the early days of August — six weeks of data behind the Q3 guide.

Q: "Can you talk about maybe intra-quarter trajectory of Vector's benefit for Unity Ad Network? Do you see pretty consistent sequential run rate increase for Unity Ads in the second quarter? And has that trend continued in July and August to date?"
— Martin Yang, Oppenheimer

A: "We are extremely pleased with the 15% sequential growth we experienced in the second quarter. I would remind investors that we did not have all components of Unity Vector fully rolled out over the course of the second quarter. So that's really exciting. We saw strength on strength. So I think each month was better than the last over the course of the second quarter. That strength has continued in the month of July and into August, which really what gives us the confidence to talk about mid-single-digit sequential growth in our overall Grow business in the third quarter."
— Jarrod Yahes, CFO

Assessment: This is the most directly bullish data point on the call. Six weeks of post-Q2 data showing continued strength is the empirical basis for the mid-single-digit Q3 Grow guide. The "strength on strength" framing inside a single quarter is what justifies the "compounding" thesis architecturally. If the Q3 print delivers, this exchange will be cited as the load-bearing forward indicator.

What They're NOT Saying

1. Full-year 2025 revenue or EBITDA guidance: Management has not provided an FY2025 guide and did not re-introduce one on this call. Implied H2 trajectory points to ~$1.77B FY revenue. The absence of FY guidance is consistent with prior quarters under this management team, but it limits Street modelability beyond the next quarter.
2. 2026 Vector contribution sizing: Management referenced "we anticipate seeing the impact [of runtime data] in 2026 and extending well into the future" but did not quantify what 2026 looks like. We expect this is deliberate — sandbagging the runway — but it leaves the bull case unanchored to a number.
3. The exact split between Unity Ad Network and "other ad products": Management said Unity Ad Network is "approximately half" of Grow and explicitly committed to not breaking out the precise split in future quarters. We respect the strategic rationale (this is competitive data) but it makes the cannibalization debate impossible for outside analysts to resolve independently.
4. Specifics on the Tencent / Scopely contract economics: The CEO declined to provide timeline or revenue contribution for the major platform partnerships, framing them as "process is not an incredibly lengthy one" but with no calendar. We treat this as appropriate confidentiality but note that the bull-case sizing of partnership revenue remains unquantified.
5. Headcount commentary: The CFO referenced "tight controls around headcount costs" but did not disclose an FTE number or a YoY headcount delta. We expect the headcount line is roughly flat-to-down — the operating leverage on the same revenue base supports that — but it would be a useful disclosure in future calls.
6. Mediation share dynamics: Management addressed the strategic question of whether mediation matters but did not provide any commentary on Unity's mediation-product (LevelPlay) market-share trajectory. If LevelPlay share is declining, the strategic answer of "we don't need to win mediation" becomes more pointed; if it is stable or growing, the answer becomes less load-bearing. The omission is telling.

Market Reaction

• Pre-print setup: Stock closed Aug 5 at $33.94, up +38.5% YTD (vs $24.51 Jan 2). Trailing 30-day +21.8% (from $27.86 on Jul 7). Options-implied move ~10% (typical for a pre-market reporter ahead of a binary print). Positioning into the print was heavy-long per typical sentiment indicators.
• Pre-market reaction (Aug 6 open): Stock opened at $37.95, gap-up of +11.8% on the headline beat. The print delivered on every measurable line — revenue +$15M above guide, EBITDA $15M above guide, EPS 38% above consensus, FCF record.
• Intraday reversal (Aug 6): Stock reversed through the day from $38.90 high to $29.75 low — a 23.6% intraday range. Closed $31.92, down -5.95% from Aug 5. Volume 46.06M shares, ~3x the trailing 30-day average. The proximate catalyst was the Q3 guide ($440–450M midpoint $445M, just below ~$447M Street consensus); the underlying cause was crowded positioning into the print after a +38% YTD run.
• T+1 (Aug 7): Open $32.78, close $33.57 (+5.16% recovery), volume 18.74M (still elevated, ~1.25x normal). Morgan Stanley raised PT $25→$40 and lifted FY26/FY27 EBITDA estimates +8% / +19%.
• Peer reaction (Aug 6–7): AppLovin (closest performance-ad-tech comparable) was modestly weaker on Aug 6 (down ~1.5%) on read-across competitive concerns. Mobile game publishers (Take-Two, Electronic Arts, Roblox) were largely flat — no obvious read-across to game-engine demand.

The intraday reversal is the most informative price-action signal of the day. A clean operating beat that gaps up +11.8% pre-market and closes -5.95% is a market response telling us positioning was crowded long, the headline-beat optionality had been priced in, and the Q3 guide was the only catalyst that mattered. Whether the guide was actually a "miss" is debatable — the high end ($450M) is +0.7% above ~$447M consensus, and management's historical sandbag pattern suggests Q3 will print at or above the high end. But the market traded the midpoint, not the high end.

The T+1 recovery to $33.57 is constructive but does not erase the message. Stock is still below the pre-print close, on volume that confirms genuine two-way debate. The Morgan Stanley PT raise to $40 (an additional 19% above the Aug 7 close) is the cleanest sell-side validation of the "generational upgrade" framing, but it is one desk; the broader Street is split between "Vector is real, buy the dip" and "Q3 guide signals a structural ceiling, sell the rally."

Street Perspective

Debate: Is Vector's sequential growth incremental or cannibalistic?

Bull view: Performance advertisers spend up to return-requirement ceilings, and Vector is raising those ceilings — empirically validated by 15% sequential growth in a quarter where the rest of Grow was flat-to-down. Management's <10% cannibalization figure is conservative; the structural ceiling for Vector growth is the size of the global performance-mobile-game-advertising market, which is multiples of Unity's current Grow revenue.

Bear view: "Approximately half" of Grow is non-Vector ad products that are declining materially YoY, and management acknowledged "resource reallocation" away from those products during Vector's rollout. That is functionally indistinguishable from cannibalization, even if customer spend is technically flat. The Q3 guide's reliance on non-Vector "stabilization" is the wager — if it doesn't stabilize, Vector is just running in place against the rest of Grow.

Our take: The bull frame is structurally correct but the data is too thin to validate empirically. Q3 is the test. If non-Vector Grow is flat or up modestly sequentially while Unity Ad Network grows 10%+, the cannibalization concern is empirically dismissed and the multiple re-rates. If non-Vector continues to decline, the bear thesis hardens. We are 60/40 on the bull side but won't pay for it ahead of the Q3 confirmation.

Debate: Does Unity need to win mediation to win in performance advertising?

Bull view: Unity's structural advantage is the engine-runtime data and first-party relationships with developers — the orchestration layer, not the auction layer. Mediation share is a tactical fight Unity doesn't need to win; the Developer Data Framework + runtime data lever creates a feature set competitors can't replicate. AppLovin's MAX is a mediation-share win, but AppLovin lacks the engine position.

Bear view: If mediation winners control the high-quality inventory and the auction-position data flow, Unity Ad Network's spend growth is fundamentally capped by what auctions mediation lets it bid on. Engine position is a content-creation moat, not an advertising-distribution moat — and they're different competitive battles. AppLovin's stock action over the past 18 months is the empirical proof that mediation-led platforms can compound faster than engine-led ones in the near term.

Our take: Both sides have a real argument. We think mediation matters less than the bear case implies because performance-advertising auctions are increasingly fragmented across in-app SDKs, web, and platform-direct integrations — there is no single auction layer that controls everything. But we also think engine position alone is insufficient — the runtime-data lever has to actually deliver in 2026 for the bull case to hold. We position Vector as the near-term lever and the data framework as the structural moat, with mediation as a tactical battle Unity will lose share in but doesn't structurally need to win.

Debate: Is the +38% YTD move discounting too much of the Vector ramp?

Bull view: At $33.94 pre-print and ~$1.75B implied FY revenue, Unity trades at ~8x EV/sales — a discount to high-quality software peers (12–18x), at a structurally improving margin profile and a multi-year product-cycle inflection. The +38% YTD move is the market starting to price the Vector thesis; if Vector continues to compound, the next leg is to $50+ on a multiple expansion to 11–12x EV/sales.

Bear view: At $33.94, the stock is up 60%+ from late 2024 lows on essentially flat trailing FY revenue. The print-day intraday reversal — pre-market +12% to close -6% — is the market telling us positioning is overextended. Anchoring to 8x EV/sales requires believing FY2026 revenue will accelerate to mid-teens YoY — which is the bull case, not the base case. If Q3 disappoints on Grow execution, the stock retests $25.

Our take: The bull case is right on the structural multiple if Vector compounds. The bear case is right on the near-term positioning. We initiate at Hold because both can be true, and the asymmetric outcome is to wait one quarter for empirical confirmation rather than chase the YTD move into a print that the market itself didn't believe.

Model Update Needed

Valuation impact: At the Aug 7 close of $33.57 and ~530M diluted shares, market cap is ~$17.8B; net debt of ~$500M brings EV to ~$18.3B. On our revised FY26 estimates of $2.05B revenue and $475M EBITDA, U trades at 8.9x EV/Sales and 38.5x EV/EBITDA. The right framework is FCF: $455M FY26 FCF estimate puts the stock at a ~25x EV/FCF multiple — reasonable for a low-double-digit-growth software business with margin expansion ahead, expensive for a no-growth ad-tech company. Which one Unity becomes is the central question. Our 12-month fair-value estimate: $36–38 (in line with current price, ~7–13% above the Aug 7 close), reflecting that the Vector thesis is now consensus-priced and additional upside requires either Q3/Q4 acceleration or a 2026 runtime-data inflection.

Thesis Scorecard Post-Earnings

Overall: Thesis confirmed on Vector inflection, Create acceleration, and operating leverage. Thesis open on cannibalization, mediation positioning, and the 2026 runtime-data lever. The bear case on valuation is the most acute near-term concern.

Action: Initiating at Hold. Vector is real; the entry is not. Wait for Q3 to deliver on the mid-single-digit Grow sequential commitment and Q4 to show the first non-Vector ad-product re-acceleration. Upgrade trigger: Q3 print with Grow at $305M+ and Unity Ad Network up 10%+ sequentially while non-Vector ad products are flat-to-up. Downgrade trigger: Q3 Grow comes in below $295M or non-Vector ad products continue to decline materially.