Summary
Adobe hit every number it promised in Q3. The numbers it didn’t promise are the ones that slowed.
- Q3 FY26 revenue
- $6.76bn (+13%)
- Non-GAAP operating margin
- 44.0% (on guide)
- Monthly active users
- 1bn+ (+20%)
- Net-new ARR
- $0.40bn (−41% YoY, like-for-like)
- RPO growth
- 8.4% (Q2: 13.1%)
- ADBE, 15 Sep 2026
- $257.76 (+17% since July)
Spider Eyes Research is long Adobe, as it was in July.
Adobe is spending growth to buy users. Free tiers come first, price rises are on hold, and management says the payback comes over 2027. Is the slowdown the cost of that strategy, or cover for a shrinking paid core?
The firm’s DCF base case is $356 a share, and it holds only if the slowdown proves to be a choice. The share price leaves room for the other answer. December can break the case. Only 2027 can confirm it.
| What the firm is watching | When | |
|---|---|---|
| Breaks the position | Evidence that paid seats are shrinking in Creative Cloud or Acrobat, or FY2027 targets pointing to growth far below the firm’s base case | December 2026 |
| Lowers the value | Q4 net-new ARR, RPO growth or the FY2027 margin target falling short of the lines set in What the firm is watching | December 2026 |
| Confirms it | Net-new ARR recovering in the first half of FY2027, free users converting to paid plans, a date for Creative Cloud price rises | March to June 2027 |
The July note, marked
Twelve calls from July, one quarter later. Q3 tested the risks harder than the strengths.
Four of July’s points moved toward the bear case, two toward the bull, four held and two went both ways. The change in how Adobe reports retired one of July’s own tests.
| July said | Now | Where |
|---|---|---|
| Growth is accelerating, not decaying | Bear ↑ | The slowdown Adobe says it chose |
| Seat compression is a risk | Bear ↑ | The slowdown Adobe says it chose |
| Pricing power is contested | Bear ↑ | The slowdown Adobe says it chose |
| Margins will drift, not collapse | Held | The cost of free |
| AI monetisation is small against the base | Held | The cost of free |
| Revenue is less seat-dependent than bears think | Bull ↑ | The quieter engine |
| Leadership vacuum is the largest swing factor | Mixed | The quieter engine |
| Canva and Apple attack whole segments | Mixed | Where the AI competition is emerging |
| Model-layer tools threaten the workflow | Bear ↑ | Where the AI competition is emerging |
| Figma threatens the design wing | Held | Where the AI competition is emerging |
| Tuck-in M&A remains open | Bull ↑ | Where the AI competition is emerging |
| The price sits well below value | Held — gap narrower | What the price leaves room for |
| New: reporting by customer group | Changes the tests | The slowdown Adobe says it chose |
The triggers July set, and what happened to each
| July’s trigger | Outcome |
|---|---|
| Seat compression: net-new Digital Media ARR slows | Tripped, measured on total net-new ARR. Adobe no longer discloses Digital Media ARR |
| Prosumer attack: Apple subscriber figures or a Canva IPO filing | Neither appeared. Canva reportedly cut its own growth outlook instead |
| Margin: non-GAAP operating margin below the “44.5% guide” | Not tripped. July took the Q2 actual for the guide; June guided Q3 lower, and Q3 landed on it |
| Leadership: CEO and permanent CFO named | CEO named internally, effective 1 December 2026. CFO still interim. Head of the creative business steps down on 27 September 2026 |
| Model layer: Gemini Omni gains traction | No adoption data. The pattern widened to Google Pics and an Anthropic design tool |
| Pricing: deferred Creative Cloud price rises return | Still deferred, no date |
| Catalyst: CEO appointment as a two-sided event | Resolved toward continuity |
| Catalyst: Topaz Labs closes | US clearance in July; close expected in Q4 FY2026 |
The slowdown Adobe says it chose
Management says the slowdown was the plan. The quarterly pattern half agrees.
ARR: Q3 stands out
Net-new ARR by quarter, US$bn, at December 2025 exchange rates, like-for-like: Semrush is included in every quarter, using the firm's estimates from Semrush's own results before the April 2026 close. Q4 FY26 is the level implied by Adobe's FY2026 ending ARR target. Source: Adobe investor datasheet, Q3 FY2026; Semrush quarterly results; firm's calculations.
Q1 is Adobe’s seasonally weakest quarter. Like-for-like, Q2 was almost level with a year earlier; Q3 fell well short. Adobe says it steered traffic to free tiers and held back price rises, by its June account in roughly equal parts. That fits ARR: a free user adds nothing until they pay, and a deferred price rise adds nothing at all.
Two cautions. Semrush’s figures before the April close are the firm’s estimates. And the FY2026 target needs Q4 to recover most of the way to last year’s level; a miss would make Q3 the start of a trend.
None of it is currency: ARR is held at fixed exchange rates. But Adobe no longer discloses Digital Media ARR, so the creative seat base cannot be seen on its own.
RPO: the explanation that doesn’t fit
Year-on-year growth, %, as reported. RPO growth is the firm's calculation from reported balances. Neither series is adjusted for Semrush before its April 2026 close; Semrush's own remaining obligations are not in its quarterly releases. Source: Adobe investor datasheets.
RPO had grown faster than ARR for a year. In Q3 it fell well below it. Management blamed the same free-tier push, and seasonality. Neither fits well. RPO is mostly multi-year enterprise contracts, where a monthly self-serve plan barely registers. The share due within twelve months held at 67%, so contracts did not get noticeably shorter. And Q4 steps up by about 10% every year, so a normal Q4 leaves growth roughly where it is.
The likelier cause is enterprise renewals: their timing, size or length. Speculation The 10-Q does not settle it: its RPO note gives no split by contract length or customer. Leaving aside what came with Semrush, deferred revenue rose only $71m in the first nine months, against $275m a year earlier, which confirms slower billing but not its cause.
The cost of free
Free users are not free. Each new dollar of revenue now costs more to serve.
- Cost to serve each extra $1 of subscription revenue
- 15¢
- Cost to serve each $1, a year earlier
- 9¢
- Non-GAAP operating margin, like-for-like
- −1.7pts YoY
Over the past year Adobe added about $680m of quarterly subscription revenue, like-for-like, and about $100m of cost to serve it. New revenue is earning a gross margin nearer 85% than the 91% on last year’s book. The 10-Q says where most of it went: hosting and data-centre costs, which include AI inference, drove 22 of the 24 points of reported growth in subscription cost of revenue. Adobe does not split out free-tier traffic, but it runs through that line. Speculation
Non-GAAP operating margin by quarter, %, like-for-like: Semrush is included in every quarter, using the firm's estimates from Semrush's own results before the April 2026 close. Q4 FY26 is Adobe's guide. Source: Adobe investor datasheet; Semrush quarterly results; firm's calculations.
The quarter was still no miss. Operating margin landed where Adobe guided, and Q1 is always the seasonal high. But the last two quarters are the lowest in the series, and Q4 is guided no higher. Below the operating line, a higher tax rate took net margin down further. The 10-Q blames smaller tax benefits abroad and a bigger tax charge on stock pay: shares granted at an average $407 vested this year at about $260, so the tax deduction falls short of the cost booked. Speculation
What is the spend buying? AI-first ARR passed $650m, still only about 2.4% of total ARR. Firefly app and credit-pack ARR grew 40% in the quarter. AI is being paid for, but not yet at a scale that moves the whole business.
The figures behind this section
| Q3 FY25 | Q3 FY26 | Change | |
|---|---|---|---|
| Subscription revenue, US$m, like-for-like | 5,902 | 6,582 | +11.5% |
| Subscription cost of revenue, US$m, like-for-like | 532 | 633 | +19.0% |
| Subscription gross margin, like-for-like | 91.0% | 90.4% | −0.6pts |
| Non-GAAP operating margin, like-for-like | 45.7% | 44.0% | −1.7pts |
| GAAP effective tax rate, as reported | 19.0% | 22.5% | +3.5pts |
| GAAP net margin, as reported | 29.6% | 27.0% | −2.6pts |
Like-for-like Q3 FY25 adds Semrush’s June to August 2025 revenue, cost of revenue and non-GAAP operating income, spread by month from its reported quarters. Semrush revenue is treated as subscription revenue. As reported, subscription revenue grew 14% and its cost 24%; operating margin fell 2.3 points.
The quieter engine
Adobe’s next chief executive comes from the part of the business that gets the least attention.
| Line | What it does | Q3 signal |
|---|---|---|
| Experience Platform and apps | Customer data and engagement; the data layer for Adobe’s marketing agents | ARR growth above 20% |
| GenStudio | Producing, adapting and approving on-brand content at scale | ARR growth above 20% |
| Experience Manager and agentic web apps | Websites, and how brands show up in AI search (Brand Visibility, built with Semrush) | ARR growth above 20%; Brand Visibility paid customers doubled in the quarter |
| CX Enterprise Coworker | An agent that runs marketing workflows across Adobe and other apps; launched in June | 1,700+ customers and early adopters |
Adobe gives thresholds, not sizes, and since the reporting change these lines share a reporting line with Creative Cloud. They also fit neither side of the “AI kills creative seats” debate, being priced more on usage and data than on seats. Speculation
“Above 20%” needs two cautions. In Q1 Adobe said “over 30%” for Experience Platform and GenStudio, so growth may be slowing. And the whole former Digital Experience business grew about 13% like-for-like, against about 11% for Digital Media: the fast lines are the growing part of a larger, slower book. Speculation
The buyers skew to banks and travel, and Publicis joins Dentsu and Stagwell as a holding-company agency win. Adobe gives no deal values and does not say which are new customers.
Named customer wins, Q2 and Q3 FY2026
| Group | Q2 FY2026 | Q3 FY2026 |
|---|---|---|
| Business Professionals & Consumers | Accenture, Datadog, KPMG, Merck, NHL, New York State Court System, The Church of Jesus Christ of Latter-day Saints, Defense Information Systems Agency, US Department of Housing and Urban Development | Amazon, Deutsche Post DHL, Honeywell, JISC, Renault, US Department of Energy Los Alamos National Laboratory |
| Creative (enterprise) | Merck, SAP, ServiceNow, Tesco, The Coca-Cola Company, Workday, Xfinity | Academy Sports, Disney, Jet2, Perficient, Premier League, Publicis, Tennis Australia, T-Mobile |
| Marketing | Dentsu, Merkle, DISA, Diriyah Company, Kaiser Foundation Hospitals, Merck Sharp & Dohme, NHL, SAP, ServiceNow, Stagwell, Stellantis, Tesco, The Coca-Cola Company | Academy Sports, Alpine Racing, BNP Paribas, Humana, IKEA, Jet2, Marriott, MSC Cruises, Publicis, Royal Bank of Canada, Vanguard, Wells Fargo |
No name repeats between the quarters. Academy Sports, Jet2 and Publicis each appear in two groups in Q3, which points to selling across Adobe rather than product by product. Source: Adobe prepared remarks, Q2 and Q3 FY2026.
Anil Chakravarthy runs this business and has led Adobe’s worldwide field operations since September 2020. He takes over as chief executive on 1 December 2026. The board has picked the enterprise side to lead. Speculation The cost is at the top of the creative business: its head, David Wadhwani, steps down on 27 September 2026, and the finance chief is still interim. The 10-Q shows the board paying to hold the team together, with retention stock and severance cover for unnamed executives until a year after the new chief executive starts.
Where the AI competition is emerging
The threat is no longer one rival. It is AI turning up inside tools people already use.
| Where | Who is moving | What Adobe has there | Pressure |
|---|---|---|---|
| Business documents and slides | Google Pics in Docs and Slides; Gemini Omni editing in Google Vids | Acrobat, Express and their free tiers | Rising |
| The first draft | Anthropic’s prompt-to-design tool; OpenAI and Midjourney image models | Firefly, with rival models running inside it | Rising |
| Marketing assets and images in the design canvas | Figma Buzz and Weave | Express, GenStudio, Firefly Services | Rising |
| Interface design | Figma Design, the standard | Project Oasis, in private testing | Steady |
| Prosumer creative apps | Canva; Apple Creator Studio | Creative Cloud, Express | Easing |
Pressure is the firm’s judgement, as of 16 September 2026. AI launches move fast. Speculation
The rising pressure lands on the users Adobe most needs to convert. Google now bundles image generation into the Workspace subscriptions it already sells. Speculation
Figma is slowing, guiding Q3 revenue only about 1% above Q2, but it is spreading onto Adobe’s ground: marketing assets and image tools inside the design canvas. Adobe’s reply, Project Oasis, is a browser design tool in private testing ahead of MAX in November, its first product aimed at Figma’s core since XD.
At the prosumer end, pressure is easing. Canva reportedly cut its growth outlook and saw its valuation marked down, citing the cost of serving AI. Speculation Apple has held its price and published no subscriber numbers.
Adobe’s defence is distribution in both directions: rival models inside Firefly, and Acrobat and Express inside ChatGPT, Claude and Microsoft Copilot. It is also still buying: Topaz Labs, about $340m in cash, cleared US review in July and is due to close this quarter.
No evidence yet shows paying customers leaving, though Adobe’s disclosures would be slow to show it.
Firefly and Figma: where the products overlap
| Job | Adobe | Figma | Overlap |
|---|---|---|---|
| Interface design and collaboration | Project Oasis, in testing (XD discontinued) | Figma Design | Figma dominant |
| Prompt to app or code | None | Make, Code Layers | None today |
| Websites | Experience Manager, Express pages | Sites (beta) | Low; different buyers |
| On-brand marketing assets at scale | Express, GenStudio, Firefly Services | Buzz (beta) | Medium and rising |
| Image generation and editing | Firefly, Photoshop generative tools | Weave tools in Figma Design | Medium |
| Vector illustration | Illustrator | Draw | Low to medium |
| Motion | After Effects, Premiere, Firefly video | Motion | Low |
| AI assistant | Creative Agent, Firefly AI Assistant | Figma Agent | Parallel |
Firm’s mapping from product descriptions, 16 September 2026. Sources: Figma Config 2026 release notes; Figma Q2 2026 results; Adobe Q3 FY2026 prepared remarks; press reports on Project Oasis.
What the price leaves room for
The firm’s base case fell, but not because of anything Adobe did.
The value moved on rates, not results
Implied value per share, US$, from the July base case to the Q3 base case. Discount rate: the 10-year US Treasury yield rose from 4.50% to 4.97%, and beta moved from July's 1.30 estimate to a sourced 1.417. Tax and capex also covers depreciation and amortisation. Source: firm's DCF model.
Lower growth and margin assumptions were more than offset by rolling the model forward and a smaller share count. The fall came from the discount rate: a higher Treasury yield, and a beta the firm now sources rather than estimates. On a five-year beta of 1.20, the base case would be about $403.
Each 0.1-point move in the 10-year US Treasury yield shifts the base case by about $5 a share. The firm does not forecast bond yields, so it uses the rate on the valuation date. At 4.97%, that is close to its highest in a decade, against a ten-year average below 3%, which makes the base case conservative on this input.
Where the price sits
Scenario values against the share price, US$ per share, 15 September 2026. Scenario settings are in the box below. These values are not price targets. Source: firm's DCF model.
- Discount rate the price implies
- 14.4% (model: 11.2%)
- Or operating margin lower every year
- −11pts
- Or revenue growth lower every year
- −4.6pts
Every temporary stress, even a severe one, leaves value above the share price. Only the structural case, with slower growth every year, thinner margins and lower long-run growth, falls below it. The price is consistent with that outcome, or with investors demanding a much higher return for the uncertainty. Speculation It does not tell you which.
The buyback does some of the work
Adobe bought back 9.5 million shares in Q3 for $2.23bn. At that pace it retires about 9.7% of its shares a year before new issuance; the diluted share count is 6.7% lower than a year ago. The cost was 92% of the quarter’s free cash flow. Stock pay of $539m in the quarter offset about a quarter of the spend. Cash fell $560m, and $1.6bn of debt now falls due within a year, which Adobe says it intends to refinance. The pace holds only while cash flow does. Speculation
Model inputs and scenario settings
| Input | Q3 model | July model |
|---|---|---|
| Risk-free rate (10-year US Treasury, 14 September 2026) | 4.97% | 4.50% |
| Beta | 1.417 (Alpha Vantage) | 1.30 (estimate) |
| Equity risk premium | 4.75% | 4.75% |
| WACC | 11.2% | 10.2% |
| Terminal growth, revenue-weighted | 2.7% | 2.7% |
| Normalised GAAP operating margin | 35.0% FY27–28, rising to 37.0% | 36.5%, rising to 37.5% |
| Scenario | Settings | Value |
|---|---|---|
| Bull | Free-tier payback +3pts growth FY28–29; pricing +1.5pts FY28–30; margin +1pt | $380 |
| Base | No payback from free tiers or pricing | $356 |
| Bear | Growth −2pts FY27–29; margin −3pts | $313 |
| Severe | Growth −3pts FY27–29; margin −5pts | $289 |
| Structural, growth only | Growth −3pts every year FY27–36 | $288 |
| Structural | Growth −3pts every year FY27–36; margin −3pts; terminal growth −1pt | $253 |
One-year betas (−0.1 to 0.8, depending on source) were rejected: Adobe’s shares have moved almost independently of the market in 2026, so a short-window beta measures that decoupling rather than lower risk. The 1.417 is a five-year measure, matching the main data providers, and reflects a period of acquisitions and this year’s shift in ARR. If those settle, a re-estimate would lower the discount rate and raise the value, upside the base case does not count. Speculation
The model assumes working capital releases cash as revenue grows; the first nine months of FY2026 went the other way, and a neutral assumption would take about $8 off the base case. Shares are 391.5 million, diluted, at 28 August 2026, the net debt date. Model revenue includes Semrush from its April 2026 close.
Bear case, then bull case
Five points each, all drawn from what this review found.
Bear case
- Net-new ARR fell 41% on the year, like-for-like, and the free-tier explanation does not fit the RPO slowdown.
- Each extra dollar of subscription revenue costs about 15¢ to serve, against 9¢ on last year’s book.
- Google is bundling AI image tools into Workspace, among the users Adobe needs to convert.
- The enterprise lines may be slowing, from “over 30%” in Q1 to “above 20%” in Q3.
- The creative business has no named leader, and price rises have no date.
Bull case
- Adobe hit every number it guided and raised its forecast for the year.
- More than 1bn monthly users, and Firefly ARR up 40% in a single quarter.
- Three enterprise lines growing ARR above 20%, under a chief executive who ran them.
- Challengers are paying the same AI bill: Canva reportedly cut its outlook, and Figma is slowing.
- Only the structural case values Adobe below the share price, and buybacks retire about 9.7% of shares a year before new issuance.
Where the position stands
Three findings changed the firm’s own reading of Adobe.
The RPO slowdown looks like enterprise renewals, not free users. Speculation About a quarter of the reported fall in operating margin was Semrush’s lower-margin business joining. And the lower valuation came from interest rates and a re-sourced beta, not from Adobe’s results.
The thesis holds, on a narrower base
- Still holds: Adobe delivered its plan, and only a structural decline puts value below the share price. Enterprise growth and the buyback support it.
- Now rests on: the slowdown proving to be a choice, not a shrinking paid core.
- First test: Q4 FY2026 results and FY2027 targets, in December 2026.
Spider Eyes Research stays long.
What the firm is watching
December can break the case. Only 2027 can confirm it.
Events in order, not to scale. Dates for Figma's Q3 results and Adobe's Q4 FY2026 results are not yet announced. Source: Adobe and Figma investor calendars; Adobe MAX; firm's selection.
| When | Event | Would count against the case | Would count for it |
|---|---|---|---|
| November 2026 | Adobe MAX; Figma Q3 | Figma Buzz and Weave winning marketing budgets | Project Oasis launched; Figma still slowing |
| December 2026 | Q4 FY2026 results and FY2027 targets | Breaks: paid-seat losses, or FY2027 organic growth nearer 5–6% than the base case’s 8.6%. Lowers value: Q4 net-new ARR below about $0.70bn; RPO growth below 8%; FY2027 non-GAAP margin target below 44% |
Q4 net-new ARR at or above the implied $0.78bn; RPO growth back towards 10%; a date for price rises |
| By the FY2027 plan | Leadership | No creative head or permanent finance chief named | Both roles filled |
| March and June 2027 | Q1 and Q2 FY2027 | Net-new ARR still falling on the year | Net-new ARR recovering, with evidence free users convert |
What this note is. It sets out a valuation of the company and the position the firm holds. Every figure in it is the arithmetic consequence of the assumptions stated beside it — not a forecast, not a target, and not a view on where the price will go — and nothing in it recommends a course of action to anyone.
Position disclosure. Spider Eyes Research and/or its principals hold a long position in Adobe Inc. (ADBE).
Important notice. This document is provided for information and discussion only. It is not investment advice, not a personal recommendation, and not an offer, invitation or inducement to engage in any investment activity. The firm is not authorised or regulated by the Financial Conduct Authority or any other regulator, and holds no licence to provide investment advice.
The firm issues no ratings, recommendations or price targets, and this note does not tell anyone to buy or sell anything. It is written once and published unchanged to everyone who receives it, takes no account of any reader’s circumstances, objectives or tolerance for risk, and creates no advisory relationship with any reader or subscriber.
Figures and estimates in this document are illustrative consequences of the assumptions stated alongside them. They are not forecasts, targets or predictions. Discounted cash flow outputs are highly sensitive to assumptions, and small changes produce large differences in result. Nothing here should be relied upon for any investment decision.
Information is drawn from public filings believed accurate at the date of writing. No representation is made as to accuracy or completeness, and no obligation is accepted to update it. The value of investments can fall as well as rise and you may get back less than you invest. Past performance is not a guide to future performance. Anyone considering an investment should carry out their own research and seek advice from a suitably qualified and regulated adviser.