Adobe Inc. (NASDAQ: ADBE) — The Cash Machine Has a Conversion Problem

Research update · August 12, 2026 · Information Technology / Application Software Reference price: $263.71 at the August 11, 2026 close Scope: Information current through August 12, 2026; fiscal years end on the Friday closest to November 30.

The analytical body in Sections 1–15 is deliberately recommendation-free. Facts, calculations, interpretations, and open questions are distinguished throughout. Unless otherwise stated, financial figures are in U.S. dollars.

⚡ Claude’s Take

The author’s independent subjective opinion; general information only, not investment advice. The analytical body below carries no position.

HOLD at $263.71; accumulate on weakness in a directional $220–$245 zone, or about 9–10x the midpoint of FY2026 non-GAAP EPS guidance. Conviction: medium. Adobe is still an exceptional cash generator: trailing free cash flow is about $10.3 billion, net debt is only about $1.0 billion, revenue and RPO are growing at double digits, and the share count is falling. Yet the June report changed the quality of the growth story. Headline ending ARR growth includes roughly $480 million acquired with Semrush. A conditional calculation that holds that acquired ARR flat through year-end reduces the maintained guide to about 8.3% for legacy Adobe; it is not company guidance. Management is asking investors to accept weaker near-term monetization while it moves more traffic into a free funnel whose conversion economics it does not disclose.

This is an abandoned-quality/value setup, not a momentum call. The stock has rebounded 36% from its June low and sits above its 21- and 50-day moving averages, but remains below its 200-day average and 62% below its five-year high. FactorsToday finds positive Quality and Value exposure but a deeply negative Momentum loading. At the current price, the headline 9.8% free-cash-flow yield is attractive, but the yield after treating stock compensation as an owner cost is nearer 7.8%. The evidence does not justify paying up for an unproven funnel, nor does it support a structural-decline short while AI-first ARR is above $500 million and growing roughly threefold.

What changes the call: clean organic ARR reacceleration and disclosed freemium conversion would turn it more bullish; a FY2027 guide in the high single digits with weak paid retention would turn it bearish. The franchise still prints cash; the new funnel has not yet earned trust.

📈 Stock Price Action — Five-Year Event Map

Adobe closed at $263.71 on August 11. Over the preceding five years it peaked at $688.37 in November 2021 and bottomed at $193.41 in June 2026; the current price is 61.7% below that peak and 36.4% above the recent low. Its trailing-52-week range is $193.41–$367.46.

The company-specific readings are strongest around the Figma announcement and the large post-earnings gaps, when Adobe materially diverged from ServiceNow, Salesforce, Autodesk, Intuit, and software ETFs. The July 2026 recovery moved with those peers, making sector mean reversion a better explanation than a new Adobe fundamental inflection. Price history is from the AZI five-year ADBE series; event dates are cross-checked to Adobe’s earnings and transaction releases.

Changes Since the June 11, 2026 Report

The prior report argued that Adobe was a quality compounder priced for its own funeral and made net-new ARR the shared bull/bear falsification variable. Six developments now matter:

1. Q2 was strong in accounting terms but weaker in organic signal. Revenue reached $6.618 billion, up 13% reported and 11% in constant currency; RPO grew 13% to $22.27 billion. Total ending ARR rose 12.5% to $27.10 billion, but included about $480 million from Semrush. Removing that amount gives a rough 10.5% Q2 organic growth proxy. More important, applying the 10.2% FY2026 target to the $25.66 billion starting base and subtracting $480 million produces an approximate 8.3% acquisition-adjusted year-end proxy. Adobe does not report either proxy, and the calculation should not be mistaken for a company KPI. It does show that acquired ARR is doing material work in the headline guide. Q2 release; Q2 transcript.

2. AI monetization became visible, but remains small against the base. AI-first ARR exceeded $500 million and roughly tripled year over year. Firefly ending ARR approached $300 million and rose about 50% sequentially. Those are substantive commercial signals, yet $500 million is under 2% of total ARR and Adobe does not disclose the overlap between AI-first and Firefly ARR, AI gross margin, or how much represents incremental rather than bundled revenue.

3. Management deliberately traded near-term ARR for a larger free funnel. Creative freemium monthly active users rose from more than 50 million to more than 90 million year over year. Acrobat plus Express MAU exceeded 850 million. Adobe deferred planned Creative Cloud pricing and line optimization and routed more traffic to free offerings; management said roughly half the second-half ARR impact came from each decision. The strategic logic is credible, but traffic is an input. Conversion, retention, credit consumption, and contribution margin are the outcomes.

4. Leadership risk increased. Shantanu Narayen remains CEO while the board searches for a successor. Adobe wants the new leader involved in FY2027 planning. CFO Dan Durn left June 15 and Steven Day became interim CFO, putting the two most important executive seats in transition at the moment Adobe is changing packaging and monetization.

5. The acquisition posture widened. Adobe closed Semrush on April 28 for $1.874 billion in cash, approximately 3.9x acquired ARR. It announced an agreement to buy Topaz Labs on June 25 for undisclosed terms, adding image/video enhancement and on-device AI. Semrush extends Adobe toward search visibility and measurement; Topaz adds model and enhancement capability. Both increase integration work during the leadership handoff.

6. The stock moved ahead of the evidence. From the June 25 low through August 11 the shares rose 36.4%, including a synchronized late-July software rally. That closes part of the valuation gap identified in June without resolving the organic ARR, free-to-paid conversion, or CEO questions.

1. Executive Summary

Adobe is best understood as three related workflow franchises rather than one creative-software product. Creative Cloud supplies professional image, video, illustration, layout, motion, and generative tools. Acrobat and Express cover document creation, editing, signing, collaboration, and lightweight content. Experience Cloud connects content, customer data, orchestration, commerce, search visibility, and measurement for enterprises. The common economic asset is not a single model or file format. It is the installed workflow: skills, files, plugins, templates, policies, integrations, collaborators, and distribution accumulated around Adobe applications.

That installed base still produces unusually strong economics. FY2025 revenue was $23.77 billion, gross margin 89.3%, GAAP operating margin 36.6%, net income $7.13 billion, and free cash flow $9.85 billion. Trailing through Q2 FY2026, revenue was about $25.20 billion, operating income $9.09 billion, and free cash flow $10.28 billion. Fiscal 2025 ROIC was about 36%, well above any reasonable cost of capital. Capital intensity is negligible: six-month FY2026 capital spending was $95 million on $13.0 billion of revenue. Recurring subscriptions, negative working-capital characteristics, and low physical investment turn accounting profit into cash.

The quality debate is therefore not whether Adobe has a profitable installed base. It does. The debate is whether generative and agentic AI expands that base, bypasses it, or does both for different customers. Professional creators and enterprise marketing teams face meaningful switching costs. A casual creator choosing among Express, Canva, ChatGPT, or a mobile application does not. Adobe’s moat is consequently wide but uneven: strongest in deep production and governed enterprise workflows, narrower at the top of the funnel and in collaborative product design.

The competitive threat has moved from hypothetical to funded. Canva ended 2025 with 260 million MAU and $3.5 billion of revenue, made the Affinity professional suite free, and is expanding through motion and marketing-intelligence acquisitions. Figma’s Q2 2026 revenue rose 48% to $370 million, net-dollar retention was 136%, and more than 80% of customers above $10,000 ARR used AI credits weekly. Foundation-model vendors and hyperscalers subsidize another layer of supply. The industry is in a capital-cycle boom: capacity arrives as engineering talent, inference subsidies, free tools, and adjacent-product expansion rather than factories.

Adobe’s response is strategically coherent. It is embedding agents into Creative Cloud, distributing Adobe capabilities through ChatGPT and Claude, planning Copilot and Gemini connections, allowing third-party models inside its applications, and charging through generative credits and enterprise outcomes. The company is attempting to become the neutral control plane where users generate with many models but refine, govern, distribute, and measure in Adobe. If successful, model proliferation strengthens Adobe’s aggregation value. If editable output becomes portable and work begins and ends inside general-purpose agents, the same proliferation weakens captivity.

The latest numbers do not settle the issue. Q2 professional subscription revenue still grew 11% in constant currency, AI-first ARR exceeded $500 million, and RPO rose 13%. Those are inconsistent with imminent franchise collapse. Yet the conditional acquisition-adjusted year-end ARR scenario is roughly 8.3%, product-level ARR disclosure has become less granular, and management has not supplied free-to-paid cohorts. The bull case is on adverse watch rather than formally broken; the bear case is stronger but not proven.

Valuation now compensates investors for a good deal of uncertainty, though less than at the June low. At $263.71 and 397.5 million shares, market capitalization is approximately $104.8 billion. Cash and short-term investments of $5.63 billion against debt and commercial paper of $6.64 billion imply enterprise value near $105.9 billion. That is roughly 10.3x trailing free cash flow and 10.8x the midpoint of FY2026 non-GAAP EPS guidance. Treating trailing stock compensation of about $2.03 billion as an owner cost reduces cash earnings to about $8.25 billion and raises the effective multiple to roughly 12.8x. The current enterprise value approximately matches a no-growth earnings-power value built on headline FCF at a 9% cost of capital, but exceeds one built on after-SBC owner earnings. The price therefore embeds modest durable growth—not collapse, but not a premium compounder outcome either.

The next two to four quarters must answer four questions: whether organic ARR returns to double digits after the deliberate H2 sacrifice; whether the 90-million-plus free Creative audience converts; whether AI credit revenue grows faster than inference expense and seat pressure; and whether a credible CEO/CFO team sets FY2027 priorities without impairing margins. Those variables matter more than MAU headlines or the July tape.

2. Business Overview

Three customer groups, one workflow thesis

Adobe reorganized reporting around customer groups, but the economic structure remains recognizable. Creative & Marketing Professionals includes the flagship Creative Cloud applications and enterprise content/experience products. Business Professionals & Consumers includes Acrobat, Express, and related document and lightweight creation products. Publishing & Advertising is small and mature. Q2 subscription revenue was $4.54 billion for Creative & Marketing Professionals, up 13% reported and 11% in constant currency, and $1.85 billion for Business Professionals & Consumers, up 16% reported and 15% in constant currency. The two major groups therefore contribute roughly 97% of subscription revenue and both still grow at double-digit constant-currency rates.

Creative Cloud is the deepest workflow. Photoshop, Illustrator, Premiere Pro, After Effects, InDesign, Lightroom, and related applications serve different stages of professional content production. A designer does not remain because the software is impossible to copy. The customer remains because fluency took years to build; files preserve editable layers and histories; plugins, fonts, templates, color systems, and review processes depend on the tools; colleagues and agencies expect compatible outputs; and the cost of a mistake on a production deadline exceeds the monthly subscription. That is demand-side captivity, reinforced by scale and breadth.

Acrobat’s advantage needs more precise language than the old shorthand of “owning PDF.” PDF has been an open ISO standard since 2008. Adobe cannot exclude rivals from reading or writing it. Acrobat’s economics come from brand trust, ubiquity, security, editing depth, signatures, enterprise deployment, integrations, and accumulated document workflows. The distinction matters: legal control of a closed standard would be a structural barrier; workflow preference around an open standard must continually earn its position. The UK Competition and Markets Authority is separately investigating early-termination fees for certain UK Adobe products, underscoring that some observed retention can reflect contract design as well as genuine product preference.

Experience Cloud is less visible to creators but potentially more captive. Enterprises join content management, audience data, campaign orchestration, commerce, experimentation, and measurement across thousands of employees and external agencies. Migration can require data mapping, identity and consent controls, integration replacement, retraining, and operational risk. Semrush broadens the loop from creating content to understanding search demand and brand visibility. Adobe’s June launches connected Semrush data to Experience Manager and added Brand Visibility for generative-search answers. The opportunity is to measure whether content appears in AI-generated responses, then create and distribute the next action inside the same stack.

Revenue model and unit economics

Subscription revenue dominates. Customers generally pay monthly or annually for product bundles, individual applications, enterprise seats, consumption credits, or contracted outcomes. This reduces revenue volatility and creates favorable cash collection. Remaining performance obligations of $22.27 billion provide visibility, though RPO is not backlog in the industrial sense and includes contracts that can vary in duration.

The model is unusually asset-light. Adobe spent only $179 million of capital expenditure in FY2025. The economically relevant reinvestment runs through the income statement: R&D, sales and marketing, cloud hosting, content, and AI inference. This is why conventional free cash flow can overstate owner economics if stock compensation is ignored. It is also why the capital cycle appears in competitors’ R&D budgets and free offerings rather than in physical capacity.

Adobe’s 89% historical gross margin is evidence of substantial value capture, but AI changes the cost structure at the margin. Subscription cost now includes inference expense. If every additional generation consumes third-party compute or royalties while pricing stays bundled, incremental gross margin can fall even as usage rises. Generative credits and outcome-based enterprise pricing are intended to align revenue with consumption. The missing disclosure is the bridge from credits consumed to recognized revenue and gross profit.

Customer concentration, geography, and cyclicality

No single customer dominates consolidated economics. The mix spans individuals, small teams, enterprises, public agencies, schools, and creative professionals. Geographic diversification reduces dependence on one economy but adds currency sensitivity, privacy regimes, and data-localization requirements. The company is not classically cyclical: subscriptions and mission-critical workflows dampen volume swings. Advertising, marketing budgets, small-business formation, and seat expansion still respond to the economy. A recession would more likely slow new seats, upgrades, and enterprise expansion than cause an abrupt revenue collapse.

The greater cyclicality today is strategic. New-user acquisition can shift quickly when free AI applications become popular. Professional migration is slower because it involves skills and process; top-of-funnel attention can move in weeks. This divergence explains why Adobe can report double-digit professional revenue growth while investors rationally worry about its future cohort mix.

Business-quality verdict: Adobe remains understandable, recurring, asset-light, and highly profitable. Its core unit of advantage is the installed workflow, not proprietary model technology or control of PDF. The counter-evidence is concentrated at the new-user edge: more alternatives, lower switching cost, less differentiated generation, and a deliberate move from paid acquisition toward freemium. Consolidated growth says the moat has not failed; reduced disclosure and missing cohort economics prevent a stronger conclusion.

3. Industry Dynamics

The addressable market is large, but the denominator is promotional

Adobe’s 2024 investor presentation estimated a 2027 total addressable market of about $293 billion: $91 billion for Creative Cloud, $47 billion for Document Cloud, and $155 billion for Experience Cloud. The company says it derives these figures from third-party research, but it does not publish a distinct serviceable market or enough methodology to eliminate overlap. Dividing current ARR by that TAM produces roughly 9%, but mixes a forward promotional estimate with a current recurring-revenue numerator and is not a defensible market-share statistic.

The useful industry insight is qualitative: Adobe participates in multiple large pools—professional creation, casual communication, documents, customer data, commerce, orchestration, and measurement. The risk is that a large TAM encourages firms to enter adjacent categories faster than Adobe can defend them. Growth in content volume does not guarantee incumbent profit growth if supply expands faster and price falls.

Three layers of AI competition

The emerging structure has three economic layers.

Foundation models and inference. OpenAI, Google, Anthropic, and other model vendors compete on quality, latency, modality, safety, and price. Training and serving models require immense capital, but performance differences decay and models can be substituted. Adobe develops Firefly models yet also offers third-party choices. That is rational customer service and an implicit admission that the model itself is not the moat.

AI-native applications and collaboration. Canva, Figma, image/video startups, coding products, and general-purpose assistants wrap models in interfaces and collaborative systems. Entry barriers to a narrow generation feature are low. Distribution, community, data, and network effects matter more. Venture capital, hyperscaler credits, and profitable cross-subsidy fund a great deal of supply.

Embedded professional and enterprise workflows. This is the layer where Adobe is strongest. The output must be editable, governed, approved, versioned, rights-cleared, distributed, and measured. Source-file history, design systems, plugins, enterprise permissions, identity data, and compliance controls create switching costs that a better demo does not immediately erase.

The industry outcome depends on where value pools. If raw generation commoditizes while workflows retain control, Adobe can aggregate models and charge for precision, governance, and outcomes. If agents in general-purpose interfaces can produce editable, interoperable assets and execute the full process, Adobe becomes a callable utility and loses customer ownership.

Capital-cycle assessment

The supply response is already visible. Figma spent $167 million on R&D and $155 million on sales and marketing in Q2 2026 on $370 million of revenue, supported by roughly $1.7 billion of cash and securities. Canva operates at far greater scale than a typical private challenger: 260 million MAU and $3.5 billion of 2025 revenue. It can make Affinity free, add Cavalry motion design, buy marketing-intelligence capability, and distribute each product through a mass-market platform. Hyperscalers subsidize models and integrate creation into productivity suites.

Classic capital-cycle reasoning predicts that unusually high returns attract capacity. Adobe’s 89% gross margin and 36% ROIC are an invitation. In software, the new capacity is human capital, inference, freemium seats, and distribution—not plant. The first economic effect is therefore pressure on entry-level prices and acquisition costs, followed later by possible pressure on professional growth if challengers build compatibility and skills ecosystems.

Adobe’s defense also uses scale. It can fund model development, license third-party models, provide indemnity and governance, distribute across hundreds of millions of users, and amortize safety and compliance work over a large base. Scale alone is not enough; a large incumbent can spend heavily and still lose the interface. Scale plus existing captivity is the relevant advantage.

Regulation and standards

Regulation cuts both ways. Enterprise customers may prefer Adobe’s provenance, commercially safe Firefly positioning, privacy controls, and indemnification. Copyright litigation or new AI rules can raise industry costs and favor scaled vendors. Conversely, Adobe depends on third-party models and content rights; liability or licensing restrictions can reduce product breadth and raise cost.

The abandoned Figma acquisition shows antitrust can prevent Adobe from buying the most important adjacent challenger. The UK CMA found high barriers in professional image/vector editing and meaningful retraining, migration, extension, and design-system costs, but also noted that open formats and multihoming mitigate lock-in. In March 2026 the CMA opened a consumer-protection investigation into early-termination fees for certain UK Adobe products. A finding against those practices could reduce contractual friction and force retention to rest more squarely on product value.

Industry verdict: demand for digital content and automation can expand quickly while returns at the application layer compress. Adobe occupies the most defensible workflow layer, but a supply boom is weakening the price umbrella around casual creation and new users. The durable path is to remain the trusted system of record and control plane across many models. The disconfirming path is an interoperable agent layer that owns intent, creation, and distribution while Adobe supplies undifferentiated tools underneath.

4. Competitive Position

The moat is wide, but not uniform

Greenwald’s test asks what prevents an equally financed entrant from reproducing the economics. Technology features alone do not. Many generation and editing capabilities can be copied; Adobe itself aggregates competitors’ models. The defensible pieces are accumulated customer habits, compatible files, trained labor, plugins, distribution, team processes, enterprise data, governance, and brand trust. Those assets create demand-side captivity and allow Adobe’s scale to matter.

The strength differs by cohort:

The table explains why both sides can cite true evidence. Professional subscription revenue can grow 11% in constant currency while the new-user moat narrows. A mass-market challenger can win attention without immediately displacing an agency’s production stack. The risk arrives through cohort replacement: if new creators learn Canva, Figma, or agent interfaces first, today’s installed-base strength can coexist with a weaker future base.

Canva: a platform-funded price attack

Canva should no longer be described only as a template tool. It ended 2025 at 260 million MAU and $3.5 billion of revenue, then made the Affinity photo, vector, and layout suite free; Affinity surpassed five million downloads. Canva AI 2.0 produces editable layered designs and adds agentic editing, brand memory, connectors, research, scheduling, and code. Cavalry adds motion design; MangoAI adds marketing intelligence.

The economic threat is cross-subsidy. Canva can offer professional capability at zero direct price to widen the funnel and monetize teams, enterprise, collaboration, printing, and adjacent services. That compresses Adobe’s entry-level price umbrella. Five million downloads do not prove paid Creative Cloud churn; downloads can be trials, duplicates, or incremental tools. The fact that Adobe’s professional customer group still grows double digits is important disconfirming evidence. The correct conclusion is erosion at the edge, not core collapse.

Figma: collaboration becomes production

Figma is a more direct network threat in product design. Q2 2026 revenue grew 48% to $370.1 million, net-dollar retention was 136%, customers above $10,000 ARR increased 34% to 15,964, and those above $100,000 increased 46% to 1,635. More than 80% of $10,000-plus customers consumed AI credits weekly. Figma is extending from design into agents, production code, governance, motion, 3D, and plugins.

Figma’s moat is different from Adobe’s. Its browser-native multiplayer canvas and design systems create team network effects. Adobe’s traditional strength is precision across image, video, document, and production formats. The competitive boundary is moving because AI reduces the need to master isolated tools and because Figma is turning approved designs into code and production assets. Adobe could not buy this network after regulatory opposition ended the proposed transaction.

Figma also illustrates the bear case against simplistic seat metrics. AI can raise output per user and expand credit revenue while reducing the number of specialized seats. Revenue may grow through usage even as the old seat model changes. Adobe needs to prove that credits, enterprise outcomes, and a larger funnel more than offset seat and price pressure.

General-purpose agents and model vendors

ChatGPT, Claude, Gemini, and Copilot can begin at user intent rather than inside a specialist application. That distribution threatens the application interface. Adobe’s response is to expose Creative Agent through ChatGPT and Claude, with Copilot and Gemini planned, while allowing external models inside Adobe products. This is both defense and concession: Adobe gains reach but shares the customer relationship.

Aggregation can still be attractive. Enterprises do not want to evaluate rights, provenance, brand rules, data handling, and output quality separately for every model. Adobe can provide a governed layer and route each task to the best model. Its approximately 50 Creative Agent tools, editable outputs, credits, and cross-application context can make the workflow more valuable than any generator.

The limiting question is pricing power. If customers view Adobe as the indispensable place where rough output becomes production-ready, credits and outcomes can support value-based pricing. If a general agent can call Adobe functions interchangeably with other services, the agent owns demand and can bargain away Adobe’s economics.

Moat-to-financial-outcome bridge

A moat claim is only useful if its failure would appear in the numbers. Adobe’s installed-workflow captivity should preserve gross margin, paid retention, pricing, and organic ARR. Weakening at the funnel should first appear as higher free MAU but lower conversion, slower individual ARR, deferred price actions, and more marketing/inference cost. Enterprise strength should appear as stable RPO, GenStudio and AEP growth, and expansion within large accounts.

Current evidence is mixed. Q2 gross growth and RPO support the moat; management’s acquisition-inclusive guide and deferral of pricing are early warnings. Disclosure is also moving in the wrong direction: Adobe no longer provides the former Creative and Document ARR series, making it harder to distinguish healthy cohorts from mix and acquisition effects.

Competitive verdict: Adobe retains a wide installed-workflow moat in professional creation, documents, and enterprise experience, supported by demand-side captivity and scale. It does not possess an equally wide moat in model technology or casual creation. The most important competitive risk is not mass overnight churn; it is losing the next generation of users while protecting the old one with bundles and buybacks.

5. Growth History and Forward Opportunities

From seat-and-price growth to funnel-and-usage growth

Adobe’s subscription transition produced a long period of predictable growth: expand the user base, raise bundle value, move customers to higher tiers, and cross-sell adjacent workflows. Revenue grew from $15.79 billion in FY2021 to $23.77 billion in FY2025 while operating margins recovered from the Figma-break-fee disruption. The next phase is harder because AI can both enlarge demand and reduce labor/seat intensity.

Management is moving the growth engine toward three inputs: free reach, AI consumption, and enterprise outcomes. Creative freemium MAU above 90 million and Acrobat plus Express MAU above 850 million create vast distribution. Generative credits convert usage into a meter. GenStudio, Experience Platform, and Semrush connect content generation to activation and measurement. The intended flywheel is: acquire free users cheaply; let agents increase content volume; convert heavy users to paid plans or credit packs; and sell enterprises governed production and measurable outcomes.

The flywheel is plausible but incompletely measured. Adobe reports MAU and selected AI ARR without conversion cohorts, customer acquisition cost, credit revenue, or AI contribution margin. More free use can improve lifetime value, or it can cannibalize paid entry tiers and increase inference cost. The H2 monetization deferral makes FY2027 the first real test.

Creative and Firefly

Firefly ending ARR approached $300 million in Q2 and grew roughly 50% sequentially. AI-first ARR across the company exceeded $500 million and roughly tripled year over year. Adobe is embedding generation and agents across Photoshop, Express, Premiere, Acrobat, and enterprise products. It also offers external models so customers need not leave the workflow for different styles or capabilities.

The opportunity is broader than replacing manual edits. Faster production can increase the number of campaigns, variants, languages, channels, and personalized assets. A creator who produces ten times as much content may consume more credits even if the employer needs fewer seats. Adobe can participate if pricing follows activity. The risk is that generation becomes nearly free and value moves to the interface, distribution channel, or proprietary enterprise data.

Firefly’s commercial-safety positioning may matter more than benchmark leadership. Enterprises care about training-data provenance, indemnity, controls, and consistent brand output. This can support adoption even when another model generates a more impressive sample. The counterargument is that major model vendors can also add indemnity and enterprise controls, while Adobe pays them for access.

Acrobat, Express, and the consumer funnel

Acrobat AI Assistant ARR roughly tripled, and Acrobat plus Express MAU exceeded 850 million. Documents are a natural agent surface: summarize, compare, answer questions, generate forms, extract tables, route signatures, and create presentations. Adobe has distribution across browsers, desktops, mobile devices, and enterprises.

The monetization challenge is abundance. Microsoft, Google, and general-purpose AI vendors can bundle similar document assistance. Adobe must attach value to trusted manipulation, permissions, signatures, workflow completion, and enterprise deployment—not merely chat with a PDF. Express has a similar challenge against Canva: its strategic value may be feeding users into higher-value Creative and enterprise products rather than producing high standalone ARPU.

Enterprise experience and Semrush

GenStudio ending ARR grew 25%; Adobe Experience Platform and native applications revenue grew more than 30%; customer-experience AI-first ARR roughly quadrupled. These management disclosures suggest enterprise AI is moving beyond pilots. Experience Cloud’s strongest proposition is a closed loop: understand audiences, create governed content, orchestrate journeys, deliver across channels, and measure outcomes.

Semrush brings search data, competitive intelligence, and generative-search visibility. Adobe paid $1.874 billion for roughly $480 million of ARR, around 3.9x, and expects approximately $280 million of FY2026 revenue because it owned the asset for only part of the year. The multiple is not demanding for a strategic software asset, but goodwill is about 67% of the purchase price and integration must produce more than an accounting ARR lift. Brand Visibility and GenStudio for Commerce are the first product connections; Adobe has not disclosed adoption, retention lift, or revenue synergy.

Topaz Labs

The pending Topaz acquisition adds image and video enhancement, specialist AI models, and on-device deployment through Neurostream. The product can strengthen professional output quality, an area where Adobe wants to remain indispensable after raw generation commoditizes. Terms are undisclosed, so return analysis is impossible. Keeping Topaz products standalone could preserve an enthusiast franchise; integration could deepen Photoshop and Premiere. The risks are paying for fast-moving model capability, product overlap, and managerial distraction.

Growth verdict

Adobe still has multiple credible growth vectors and enough distribution to make even modest conversion economically meaningful. The best evidence is not management’s TAM; it is double-digit subscription growth, $22.27 billion of RPO, expanding AI ARR, and enterprise application growth. The negative evidence is a conditional acquisition-adjusted year-end ARR scenario near 8.3%, deferred monetization, and missing cohort disclosure.

The June bull thesis is therefore on adverse watch. Its literal falsification required negative net-new ARR or FY2027 growth guidance falling toward high single digits; neither event has occurred because FY2027 guidance is not yet issued and Adobe does not disclose the old net-new measure. Economically, however, the current proxy has already reached the warning zone. The June bear thesis required organic reacceleration plus AI ARR outgrowing seat pressure while total ARR held. AI ARR accelerated, but organic reacceleration is not visible; the bear has not been falsified.

6. Financial Quality

Multi-year operating record

Adobe’s financial record is the strongest part of the case. The following table rebuilds FY2022–FY2025 from the filed statements and derives trailing twelve months as FY2025 plus H1 FY2026 less H1 FY2025. Free cash flow is cash from operations less purchases of property and equipment. “FCF less SBC” is a conservative owner-earnings proxy, not a GAAP measure and not literally cash paid for awards in the period.

Revenue compounded at 10.8% from FY2021 through FY2025. Gross margin expanded 110 basis points from 88.2% to 89.3%. The FY2024 operating-margin trough needs normalization: it includes the $1.0 billion Figma termination fee. Adding that cost back directionally lifts FY2024 operating margin from 31.3% to about 36.0% and FCF from $7.87 billion to about $8.87 billion. FY2025’s rebound therefore combines real leverage with removal of a one-time failure cost.

The cash conversion is genuine. TTM cash from operations of $10.48 billion exceeds net income of $7.23 billion, while capital spending is only $201 million. Deferred subscription collections help working capital. The resulting 40.8% headline FCF margin is rare even among scaled software companies.

The qualification is stock compensation. TTM SBC of $2.03 billion equals 8.1% of revenue and 19.8% of headline FCF. Subtracting it yields $8.25 billion and a still-exceptional 32.7% margin. Investors should not subtract SBC from FCF and then separately assume full dilution without understanding the double count, but neither should they treat buybacks needed to offset awards as optional capital returns. The report uses both headline and after-SBC measures to bound economics.

Q2 and guidance quality

Q2 revenue rose 12.7% to $6.618 billion; subscription revenue rose 13.7% to $6.416 billion. Gross margin remained 89.2%. GAAP operating margin fell 210 basis points to 33.8%, reflecting a $70 million Publishing & Advertising goodwill impairment, loss contingencies, and acquisition costs. H1 cash flow improved: CFO was $5.123 billion, capex $95 million, and FCF $5.028 billion versus $4.600 billion a year earlier. After $1.045 billion of SBC, the H1 owner proxy was $3.983 billion.

Subscription cost grew 16% in Q2 and 13% in H1, compared with subscription revenue growth of 14% and 13%. The filing identifies hosting and data centers, including AI inference, as components. Gross margin remained stable, so there is no evidence yet of material AI margin damage. The cost line nevertheless deserves attention because a bundled product can hide rising usage cost until renewals.

Adobe raised the FY2026 revenue range from $25.90–$26.10 billion to $26.50–$26.60 billion after Semrush. It raised non-GAAP EPS from $23.30–$23.50 to $24.35–$24.45 and lowered expected diluted shares from about 403 million to 399 million. GAAP EPS midpoint, however, declined by $0.05 to $17.95, and the total ARR growth objective stayed 10.2%. Revenue and adjusted EPS growth therefore benefited from acquisition and repurchases; ARR did not receive an organic upgrade.

Q3 guidance calls for $6.67–$6.72 billion of revenue, $4.40–$4.45 GAAP EPS, $6.05–$6.10 non-GAAP EPS, and roughly 395 million shares. The falling share count is doing measurable work in EPS. That is legitimate if repurchases occur below intrinsic value and remain funded by owner earnings; it becomes financial engineering if it masks weak per-enterprise growth or requires persistent leverage.

Returns on capital and earnings power

ROIC.ai calculates FY2025 ROIC at about 36.3%, return on capital at 38.2%, ROE at 55.4%, and ROA at 23.9%. Exact definitions differ across data providers, but the conclusion does not: Adobe earns far more than its cost of capital. Tangible invested capital is low because internally developed software, skills, brand, and customer relationships are expensed. High returns are real, though balance-sheet ROIC mechanically benefits from buybacks that reduce equity.

The earnings-power question is what level of cash is maintainable without growth investment. Headline TTM FCF of $10.28 billion includes all expensed R&D and selling costs, so it is after substantial reinvestment. Conversely, deferred revenue, acquisition timing, tax items, and working capital can make one year generous. The after-SBC $8.25 billion figure is the more conservative anchor. It still implies an owner-earnings margin above 32% and ample self-funding ability.

Accounting and disclosure

The 60-month SEC census found five 10-Ks, fifteen 10-Qs, forty-seven 8-Ks, hundreds of ownership filings, and no late 10-K/10-Q, restatement, or non-reliance event. The latest 10-Q says disclosure controls were effective and internal control had no material change. That supports basic reporting reliability.

The greater concern is decision-useful granularity rather than accounting integrity. Adobe now reports one segment and no longer gives the former Creative versus Document ARR series. Semrush results are not separately material enough for pro forma disclosure. AI-first ARR, Firefly ARR, MAU, and selected application-growth figures are management-defined. They help track strategy but do not reconcile cleanly to recognized revenue, gross profit, retention, or cash.

Financial-quality verdict: Adobe remains a high-return, low-capex, cash-converting franchise. Stable gross margin and H1 cash growth argue against present economic deterioration. The disconfirming evidence is that adjusted EPS and per-share growth look stronger than organic enterprise growth because acquisitions, exclusions, and repurchases contribute. The appropriate underwriting base is both $10.28 billion headline FCF and $8.25 billion after-SBC owner earnings—not adjusted EPS alone.

7. Capital Allocation

Repurchases: effective, aggressive, and no longer costless

Adobe does not pay a dividend. It directs most excess cash to repurchases and selected acquisitions. Cash buybacks were $6.55 billion, $4.40 billion, $9.50 billion, and $11.28 billion in FY2022 through FY2025. The last two years exceeded headline FCF by $1.63 billion and $1.43 billion, respectively. FY2025 repurchases were 1.43x FCF less SBC.

The share result is real. Cover-page shares outstanding fell from 457.8 million in January 2023 to 452.0 million in January 2024, 435.3 million in January 2025, 410.5 million in January 2026, and 397.5 million in June 2026—a 13.2% decline. H1 FY2026 repurchases of $4.59 billion retired 16.6 million shares at an implied $276.39 each, versus $6.75 billion for 15.6 million shares at $432.69 in the prior-year period. Less cash retired more shares; price discipline improved materially.

Adobe now has overlapping authorizations: $25 billion approved in March 2024 through March 2028 and another $25 billion approved in April 2026 through April 2030. About $26.78 billion remained at Q2. Authorization is not cash. It equals more than three years of after-SBC owner earnings and should not be treated as a distributable asset.

The balance sheet shows the constraint. Cash plus short-term investments were $5.626 billion; debt and commercial paper were $6.645 billion, producing $1.019 billion of net debt. Six months earlier the comparable position was $385 million net cash. The $1.404 billion swing reflects acquisitions and repurchases. Adobe remains conservatively financed: $6.15 billion face value of senior notes have no financial covenants, the $1.5 billion revolver is undrawn, and $7.15 billion of deferred revenue helps fund operations. Yet buybacks above owner earnings cannot continue indefinitely alongside acquisitions without leverage rising.

Semrush: a reasonable entry multiple with a high synergy burden

The $1.874 billion Semrush purchase price equals 3.90x the disclosed $480 million acquired ARR, or 3.36x net of $262 million acquired cash. That is not obviously excessive for a growing strategic software platform. The accounting allocation reveals what Adobe bought: $415 million of technology, $107 million of customer relationships, $60 million of trademarks, and $1.251 billion of non-deductible goodwill, plus other net liabilities. Goodwill is 66.8% of consideration.

High goodwill is normal for software but makes the investment case depend on retention, integration, and cross-sell rather than recoverable assets. Adobe wants to connect search demand and brand visibility to content production and commerce. The first integrations are live, but standalone revenue, margins, retention, and synergies are undisclosed. Q2 included only about one month and $40 million of Semrush subscription revenue. Investors should not infer success from the acquired ARR appearing in Adobe’s total.

Topaz and the acquisition pattern

Topaz fits the strategic desire to own the refinement layer: image/video enhancement, specialist models, and on-device inference can strengthen output after generation. No price or financial profile is disclosed. That prevents return-on-investment analysis. The timing raises execution risk because Adobe is integrating Semrush while choosing a CEO and operating with an interim CFO.

The failed $20 billion Figma proposal remains an important capital-allocation scar. Adobe incurred a $1 billion termination fee without acquiring the asset. The episode shows both strategic awareness—the company recognized collaborative design as important—and willingness to pay an extraordinary price for growth. Semrush’s lower multiple and H1 buyback prices show better discipline, but one year does not erase that history.

Compensation, incentives, and insider evidence

FY2025 CEO compensation was $51.2 million, including $45.4 million of stock awards; the disclosed CEO pay ratio was 217:1. FY2026 performance shares are half relative TSR against the Nasdaq-100 and half total Adobe ending ARR growth, with 0–200% payout. The ARR metric can be adjusted for material M&A and currency. The annual cash plan gates at 95% of revenue and non-GAAP EPS.

The design rewards market performance and growth, but omits ROIC, FCF, after-SBC owner earnings, and explicit organic ARR. That omission matters when an acquisition can lift reported ARR. The 2026 say-on-pay vote received 148.8 million votes for and 145.0 million against—only 50.65% support among votes cast for or against. This is a clear governance warning, not a technical footnote.

The 24-month Form 4 sweep is mixed. Director David Ricks bought 10,000 shares at $194.51 on June 25, 2026, after buying 2,250 at $443.98 in January 2025. The countercyclical $1.95 million June purchase is a meaningful positive signal. Narayen sold 75,000 shares around $244 in April 2026 in a transaction described primarily as tax/estate planning but not marked as a 10b5-1 trade. Durn sold shares before his departure. Tax-withholding dispositions and award grants are not discretionary trades and were excluded from this interpretation.

Capital-allocation verdict: Adobe has created real per-share accretion and recently bought shares at much better prices. It remains self-funding and lightly levered. The program is nevertheless aggressive relative to after-SBC cash, and the balance sheet has moved from net cash to net debt. Semrush’s entry multiple is sensible but its 67% goodwill makes synergy delivery decisive; Topaz cannot be judged without terms. Governance deserves a discount because incentives omit owner earnings and shareholders nearly rejected pay.

8. Changes and Headwinds — Last Two Years

Growth disclosure and strategic reset

Two years ago the central concern was slowing Digital Media net-new ARR. Adobe now presents one total-ARR growth figure and customer-group subscription revenue. The new structure reflects product convergence, but it also removes the product-level series investors used to test Creative and Document health. At the same time, AI metrics have multiplied: AI-first ARR, Firefly ARR, MAU, agent use, and selected enterprise growth. More strategic commentary accompanies less direct reconciliation to core cohorts.

The FY2026 reset is explicit. Adobe postponed Creative pricing and line optimization and sent more traffic to freemium. Management estimates roughly half the second-half ARR impact from each action. This is not merely macro weakness; it is an internal decision to sacrifice current monetization for future conversion. The decision may maximize lifetime value. It also creates an easy narrative shield: near-term softness becomes intentional while future cohorts remain undisclosed.

Management transition

Narayen announced that he would step down once a successor is selected. Adobe wants that person engaged in FY2027 planning, but no appointment was disclosed by the report date. Durn’s departure on June 15 and Day’s interim appointment add finance leadership uncertainty. A July retention arrangement for Chief Legal Officer Louise Pentland provides salary, bonus continuation, and partial equity acceleration under specified termination conditions during the transition.

Succession can be positive if a new leader simplifies the portfolio, improves product disclosure, and aligns monetization with AI usage. It can also produce strategy churn, executive departures, and reset guidance. The CEO and CFO transitions overlap with Semrush integration, Topaz closing, freemium conversion, and new buyback authority. The operational burden is unusually high.

Competition moved closer to the core

Canva’s free Affinity suite, motion expansion, and editable AI outputs move it toward professional creation. Figma’s growth and AI-credit adoption move it beyond interface design. General-purpose assistants can call Adobe tools while owning the user’s starting intent. The threat is no longer a list of small feature competitors; it is an ecosystem contest among well-capitalized platforms.

Adobe is responding with model choice, embedded agents, distribution through third-party assistants, enterprise governance, and acquisitions. That response makes the company more relevant across the stack, but also exposes it to licensing, inference, privacy, and intellectual-property costs controlled by others.

Regulatory and legal friction

The CMA’s March 2026 consumer-protection investigation examines cancellation terms and early-termination fees for certain UK products. No infringement finding has been made. The economic relevance is straightforward: if contractual friction contributes to retention, a remedy could increase churn or reduce realized lifetime value. The failed Figma transaction separately shows that regulators can limit Adobe’s ability to buy adjacent network effects.

AI introduces content-rights and provenance risk. Adobe’s commercially safe positioning can attract enterprise customers, but training-data claims, third-party model licensing, and generated-output liability can create costs. The 10-Q explicitly warns that access to models and data, privacy, IP liability, and fast-moving competition may impair monetization.

What improved

The picture is not uniformly negative. AI-first ARR accelerated above $500 million, Firefly approached $300 million, enterprise applications grew, RPO remained double digit, gross margin held, H1 FCF increased, and repurchases occurred at far lower prices. Adobe broadened distribution rather than insisting on closed proprietary models. Semrush entered at a modest revenue multiple compared with the abandoned Figma price.

Changes verdict: operating quality held while evidence quality worsened. Adobe is monetizing AI, but its strategic pivot moved the decisive proof into FY2027 and reduced product-level visibility. Leadership, competition, and integration demands rose together. Investors should credit the cash and adoption already visible, but refuse to count free users or acquired ARR twice.

9. Risk Analysis

The most probable adverse path is not catastrophic loss. It is a slow combination of weaker new-user share, lower price realization, higher AI cost, and continued buybacks that preserves EPS longer than enterprise value growth. Adobe’s low capex, recurring revenue, liquidity, and deep installed base make insolvency or total loss remote under ordinary stress.

The highest-impact idiosyncratic risk is a failed monetization transition. If Adobe teaches a much larger audience to use free products but cannot convert it, the company bears inference and development cost while rivals capture paid team workflows. Because Adobe does not publish cohorts, the market may discover the failure only through ARR guidance and margin after several quarters.

The most important mitigant is existing professional and enterprise captivity. Customers do not replace years of files, skills, integrations, governance, and review practices because one rival feature is better. Revenue, RPO, and cash can remain resilient while Adobe experiments. That time is valuable, but it is not unlimited.

10. Valuation Discussion (Embedded Expectations)

Filing-reconciled starting point

At the August 11 close of $263.71 and the June 11 filed share count of 397.5 million, equity value is $104.825 billion. Adding $1.019 billion of net debt gives enterprise value of $105.844 billion. Using the filed spot share count avoids the common error of mixing a weighted diluted count with a point-in-time price.

No single multiple is definitive. EV/revenue understates Adobe’s margin advantage. Non-GAAP P/E excludes an economically relevant compensation cost. Headline FCF benefits from issuing equity to employees. The range between 10.2x headline FCF and 12.7x after-SBC owner earnings best captures the debate.

Peer comparison

The comp set combines mature profitable workflow suites with direct design challengers. August 11 prices were applied to the latest filed spot shares and net debt/cash; TTM denominators run through each company’s latest quarter. Growth periods are not perfectly matched. This is a directional comparison, not a valuation formula.

Adobe carries the lowest EV/EBIT and owner-cash-flow multiple among profitable suite peers while posting the highest TTM GAAP operating margin in this set. Faster Autodesk, ServiceNow, and Figma growth deserves a premium. The more striking comparison is with similarly mature Salesforce and Intuit, where Adobe still trades lower. That discount is consistent with a shorter expected moat duration, weaker new-user economics, and more direct generative-AI substitution. It is not, by itself, evidence of mispricing.

What is embedded

A single-stage Gordon inversion is a useful first diagnostic. Enterprise value equals 12.83x TTM after-SBC cash flow, or a 7.79% enterprise owner-cash-flow yield. At a 9% required return, that implies only about 1.2% perpetual growth if today’s owner margin is immediately sustainable forever. The simplification hides transition years and should not be read as a forecast.

A five-year reverse DCF shows the margin-growth trade more honestly. Starting from $25.198 billion of revenue and a 32.7% owner margin, with 2.5% terminal growth, the constant revenue CAGR required to cover current EV is:

The current price therefore underwrites low-single-digit long-run growth if after-SBC margin resets toward 25%–28%, or modest contraction if today’s 32.7% margin persists. This is the central valuation insight. The market is not pricing an immediate revenue collapse. It is pricing a joint possibility that growth slows and some of Adobe’s exceptional cash margin is reinvested or competed away.

Explicit scenarios

Each case starts from TTM revenue and uses total-company growth, so the first forecast year includes acquisition annualization. Owner margin is FCF less SBC divided by revenue. Net share change is shown as an ownership sensitivity and is not credited in the enterprise DCF, avoiding double counting. “DCF/current EV” is a coverage ratio—not a price target, fair-value range, or recommendation.

The bear needs two things to go wrong simultaneously. AI-native/free tools must impair new-user acquisition and pricing enough to flatten revenue, and inference/SBC/reinvestment must pull owner margin toward 25%. Flat revenue alone does not produce the modeled shortfall if current cash margins survive.

The base assumes installed professional, document, and enterprise captivity survives while total growth settles below the current rate. AI credits and Semrush contribute, but there is no return to historical double-digit organic compounding. Owner margin rises only modestly and repurchases slow. The fragile assumption is conversion: free and AI usage must become durable paid revenue before deferred pricing and low-end competition damage cohort mix.

The bull requires Adobe to become a model-neutral workflow control plane. AI-first ARR scales well beyond its current share, Creative/Acrobat funnels convert without professional churn, and enterprise content-data-measurement integration supports 34%–38% owner margins despite inference. Buybacks must be funded from owner earnings rather than more leverage.

Earnings power value and replacement economics

At a 9% capitalization rate, no-growth headline FCF supports approximately $114 billion of operating value, close to current EV. No-growth after-SBC cash supports about $92 billion. These are steady-state diagnostics, not liquidation values. They say the current enterprise price roughly capitalizes headline earnings power but requires some growth to justify the owner-cost-adjusted base.

Book equity of $11.52 billion is not a sensible replacement-cost floor. Adobe’s economic assets—software, skills ecosystem, brand, files, workflows, customer relationships, and distribution—are mostly absent from the balance sheet. Reproducing the features might cost far less than current EV; reproducing the installed customer system would cost far more than book. The gap between tangible book and enterprise value is the franchise value that organic ARR and margin must defend.

What the market may have right or wrong

The market correctly recognizes a creative-software supply boom, low switching cost for casual users, Canva/Figma/model-platform pressure, opaque product ARR, real inference and SBC costs, leadership transition, and less balance-sheet room after Semrush and buybacks. A lower multiple than faster peers is rational.

The market may be applying low-end disruption too uniformly across Acrobat, deeply learned professional workflows, and embedded enterprise systems. A 7.9% owner yield requires little growth if margin endures, while revenue, RPO, the acquisition-subtracted Q2 ARR calculation, and AI-first ARR do not show franchise-wide contraction. Conversely, valuation is not protected if interoperable agents erode workflow captivity and force both 25% owner margins and near-zero growth. That joint outcome—not AI adoption alone—is the decisive downside condition.

11. Variant Perception

What the tape appears to believe

The share price and factor profile imply neither imminent collapse nor confidence in a durable premium compounder. Adobe trades as abandoned quality: FactorsToday’s all-factor model assigns positive Quality (+0.573) and Value (+0.335) loadings, a deeply negative Momentum loading (-0.725), and strong exposure to the battered Software factor. The model explains about 57% of historical variance, leaving roughly 43% to company-specific developments. One-year return was about -21% through August 11 even after a 36% recovery from the June low.

At roughly 10.3x headline FCF and 12.8x after-SBC owner earnings, the market appears to underwrite low-to-mid-single-digit durable cash growth rather than the old double-digit compounder. That is reasonable if Creative pricing weakens, new-user share leaks, and AI adds cost before revenue. It is too pessimistic if Adobe converts its free audience, holds professional retention, and turns model proliferation into credit and enterprise workflow revenue.

Consensus-style case

The middle view accepts that Adobe remains indispensable to many professionals but assumes AI permanently lowers the quality and predictability of growth. Revenue can expand at mid-to-high single digits, margins remain excellent, and buybacks support per-share growth, yet the business no longer deserves the premium once attached to predictable seat-and-price compounding. Semrush and Topaz add capabilities but do not transform the trajectory. This interpretation fits current valuation and the acquisition-inclusive guide.

Bull variant

The constructive variant is not that Firefly is the best foundation model. It is that model competition makes generation cheap while increasing the value of Adobe’s editing, governance, provenance, distribution, and measurement layer. Ninety million-plus free Creative users become a low-cost acquisition pool; credits monetize heavy usage; professional cohorts retain; and enterprise customers buy end-to-end outcomes. Adobe grows cash in the high single digits or better while retiring shares at modest multiples.

Evidence supporting this case includes Q2 double-digit subscription growth, 13% RPO growth, AI-first ARR above $500 million, Firefly’s 50% sequential growth, stable gross margin, and strong enterprise-application growth. The fragile assumption is conversion. MAU without paid cohort and margin data is reach, not economics.

Bear variant

The negative variant is cohort decay hidden by the installed base and capital returns. Existing professionals renew, but new creators learn Canva, Figma, and general-purpose agents. Free Adobe usage rises while paid individual ARR and pricing slow. Enterprise products face bundling from hyperscalers and specialized platforms. Inference, licensing, and sales costs rise. Acquired ARR and share retirement keep headline growth respectable for a time, but enterprise value compounds slowly or declines.

Evidence supporting this case includes the conditional 8.3% legacy-Adobe year-end ARR scenario, deferred pricing, reduced product disclosure, strong Canva/Figma growth, and buybacks above owner earnings. Evidence against collapse is equally important: 89% gross margin, double-digit subscription and RPO growth, substantial AI ARR, and very low leverage. The fragile bear assumption is that challenger adoption eventually overcomes professional and enterprise switching costs.

Variant verdict

The differentiated conclusion is narrower than either extreme. The market is right to remove Adobe’s old certainty premium and wrong to treat every creative layer as equally vulnerable. Generation is competitive; installed production and governance are captive. The shares will be determined by the conversion bridge between those truths. Organic ARR, paid retention, credits, and after-inference margin—not raw AI usage—are the decisive proof.

12. Fact vs. Interpretation

13. Open Questions

1. What is total Adobe ARR growth excluding Semrush on a consistent currency and acquisition basis, and what is net-new ARR by major customer group?
2. What proportion of Creative freemium cohorts converts to paid plans after 30, 90, 180, and 365 days, at what acquisition and inference cost?
3. How much AI-first ARR is incremental versus bundled, and how much overlaps Firefly ARR? What are AI credits’ revenue and gross margin?
4. Are professional Creative retention and expansion stable while casual/new-user share shifts, or is weakness already entering installed cohorts?
5. Does distributing Creative Agent through general-purpose assistants create incremental Adobe demand or transfer customer ownership and bargaining power to those assistants?
6. What portion of content production begins in Adobe products, and how is that changing by cohort?
7. What are Semrush’s standalone growth, retention, margin, and revenue synergies after integration? How will M&A be adjusted in compensation metrics?
8. What consideration and return assumptions underlie Topaz? Will its products remain standalone, and how will enhancement capability be monetized?
9. Who becomes CEO and permanent CFO, and will the new team preserve the funnel strategy, margin structure, and repurchase pace?
10. Will FY2027 guidance return organic ARR to double digits, or formalize the high-single-digit warning now visible in the conditional bridge?
11. How much of subscription cost growth is AI inference and third-party licensing, and can credits offset it without weakening the value proposition?
12. What remedy, if any, results from the CMA cancellation-fee investigation, and how much retention reflects contractual friction?
13. Will the April $25 billion authorization be paced to after-SBC owner earnings, or will net debt rise further?
14. Why did Adobe move to one segment and less granular ARR disclosure during a strategic transition, and what replacement KPIs will it provide?

14. What Must Be True

Constructive case

The constructive case requires four linked outcomes. First, the installed professional and enterprise base must remain captive enough to sustain pricing and retention. Second, the enlarged free funnel must convert to subscriptions or credits at attractive lifetime value after inference cost. Third, AI usage must be additive: credit and outcome revenue must exceed any seat compression and model expense. Fourth, management must constrain buybacks and M&A to owner earnings while completing succession.

Observable scorecard:

Falsification test: the constructive case is broken if acquisition-adjusted net-new ARR turns negative, or if FY2027 guidance formalizes high-single-digit-or-lower growth without credible cohort evidence. It is also broken if paid professional retention weakens while free MAU rises, because that would show cannibalization rather than funnel investment. Current status: adverse watch, not formally broken. The conditional 8.3% year-end legacy scenario is already in the economic warning zone, but it is not company guidance and assumes no growth in Semrush ARR after closing.

Negative case

The negative case requires gradual erosion rather than instant obsolescence. New users form habits in Canva, Figma, and agent interfaces; Adobe protects existing cohorts but loses share at replacement. Bundled AI caps price, inference raises cost, and enterprise aggregation does not fully offset Creative softness. Acquisitions and buybacks make reported and per-share outcomes look better than organic enterprise growth.

Observable scorecard:

Falsification test: the negative case is broken if Adobe discloses clean organic ARR reacceleration while AI-first ARR outgrows seat pressure, total ARR holds without acquisition support, and free-to-paid cohorts earn attractive contribution margins. Current status: not falsified. AI monetization accelerated, but there is no disclosed organic reacceleration; headline ARR includes Semrush.

Shared decision rule

Both cases turn on one variable: acquisition-adjusted net-new ARR, supported by conversion and margin evidence. Revenue can lag competitive change because subscriptions renew. MAU can lead revenue but can also be vanity. EPS can be supported by exclusions and buybacks. Organic ARR connects demand, conversion, pricing, and retention; gross margin and after-SBC FCF show whether that growth creates owner value. Until Adobe restores that bridge, certainty in either direction is not earned.

15. Source Notes

The analysis prioritizes Adobe’s 10-K, Q2 10-Q, 8-K exhibits, proxy, and corrected Q2 transcript; Adobe product releases; the UK CMA and ISO; and Figma and Canva primary disclosures. ROIC.ai supplies ready-computed cross-checks that were reconciled to the filings. AZI supplies the five-year price series; FactorsToday supplies empirical factor exposures. Secondary reporting is used only for contemporaneous market interpretation or short-interest context. Appendix B provides the dated source inventory and limitations.

APPENDIX A — Standard Diligence Questionnaire

General

What thoughtful questions are other investors asking? The highest-quality questions focus less on whether Adobe has AI products and more on whether those products change economic outcomes. Is AI-first ARR incremental, bundled, or overlapping with Firefly ARR? What percentage of Creative users who enter free plans become paying users, how long does conversion take, and what inference cost is incurred before conversion? Is the acquisition-adjusted ARR slowdown deliberate and temporary, or an early signal of lost pricing power? Can Adobe own the workflow while ChatGPT, Claude, Gemini, or Copilot owns the user interface? Why did Adobe stop disclosing Creative and Document ARR just as these questions became material? How will a new CEO and CFO allocate capital after years of unusually large buybacks?

Another thoughtful line of inquiry separates cohorts. Professional editors, enterprise marketers, knowledge workers, and casual creators do not have the same switching costs. A consolidated revenue result can conceal a weakening new-user funnel alongside a stable installed base. Investors should ask for paid retention, app-level ARR, free-to-paid conversion, credit revenue, and AI contribution margin by major customer group.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Adobe is not at a classic volume-cycle peak. Revenue is subscription-heavy, diversified, and still growing at double digits. TTM gross margin of 89.4% and owner-earnings margin of 32.7% are structurally high, but not obviously inflated by a temporary pricing spike. The most important cycle is technological and competitive: generative AI is attracting supply and changing how customers acquire and use creative software.

External environment or internal actions? Both. External pressure comes from model vendors, Canva, Figma, hyperscaler bundles, and customer experiments with general-purpose agents. Internal actions amplify near-term uncertainty: Adobe deferred Creative pricing/line changes, routed more traffic to freemium, closed Semrush, agreed to buy Topaz, began a CEO search, and appointed an interim CFO. Management says roughly half the H2 ARR impact comes from deferred pricing and half from the funnel shift.

How stable are revenues? RPO of $22.27 billion, recurring subscriptions, broad customer diversification, and installed workflows support stability. Economic weakness would more likely slow seats, expansion, and marketing spend than abruptly collapse revenue. The new uncertainty is conversion: free usage and AI consumption can rise without paid ARR keeping pace.

Outlook for products and services? Professional Creative and enterprise Experience products should remain resilient because of depth and integration. Acrobat has wide distribution but faces bundled document AI. Express and entry-level Creative are most exposed to free alternatives. Firefly and agents can raise content volume; credits must convert activity into revenue. Semrush can connect search intelligence to content and measurement; Topaz can strengthen refinement quality. Neither transaction’s synergy is proven.

How large is the market? Adobe estimates a 2027 TAM of about $293 billion across Creative, Document, and Experience. This is a broad management estimate with possible overlap and no separately disclosed serviceable market. It is useful as evidence of adjacency, not a market-share denominator. Demand is global and likely growing as content volume and automation expand, while supply is growing even faster in some layers.

Business Quality & Competitive Moat

Is the industry becoming more competitive? Yes. Capital is entering foundation models, AI-native applications, collaboration, free professional tools, and productivity bundles. Canva can cross-subsidize Affinity and other products. Figma spends heavily to expand from design into code and production. Model vendors distribute creation through general assistants. Adobe is responding with model aggregation, agents, credits, enterprise governance, and acquisitions.

How profitable is Adobe? FY2025 ROIC was approximately 36.3%, ROE 55.4%, gross margin 89.3%, GAAP operating margin 36.6%, and FCF margin 41.4%. TTM FCF was $10.28 billion; after subtracting $2.03 billion of SBC, the owner proxy was $8.25 billion. These economics are far above a reasonable cost of capital.

How profitable is the industry? Scaled workflow vendors can achieve high gross margins, but challengers often reinvest heavily. Figma’s Q2 GAAP gross margin was 84% while R&D and sales/marketing together approached its quarterly revenue. Foundation-model economics are capital-intensive and fast-changing. Adobe’s high returns attract entry; the sustainability of industry profit depends on captivity and distribution rather than feature scarcity.

What are the barriers to entry? A narrow generation or editing feature has a low barrier. A professional workflow has higher barriers: learned skills, editable source files, plugins, fonts, design systems, review processes, enterprise permissions, integrations, data, security, and brand trust. Scale matters because Adobe can amortize safety, model access, hosting, and enterprise sales across a huge base. The combination of captivity and scale—not proprietary technology alone—is the moat.

Can the business be understood? Yes at the economic level. Adobe sells recurring access and usage across creation, documents, and enterprise experience. The difficulty is measurement: one reporting segment, acquisition-inclusive ARR, and management-defined AI KPIs obscure product-level trends.

Can low-cost foreign labor undermine it? Labor arbitrage can lower the cost of implementation, support, and content production, but cannot directly reproduce Adobe’s installed workflows or distribution. AI is the more important substitute because it reduces the skill and time required for creative work globally. That can lower seat demand while increasing total output.

Do brands matter? Yes. Adobe and Acrobat communicate professional compatibility, reliability, security, and trust. Brand matters most when work must be production-ready or legally sensitive. It matters less for casual generation where users can trial alternatives immediately.

What is the nature of competition? Competition occurs on model quality, workflow depth, collaboration, price, distribution, interoperability, trust, and measurable outcomes. Canva competes through ease and price; Figma through collaboration and network effects; Microsoft/Google through bundling; model vendors through intent-level interfaces; specialists through superior narrow tasks.

What are customer switching costs? High for enterprise marketing stacks and team production; medium to high for professional individuals; medium for enterprise Acrobat deployments; low for casual creators and new users. Multihoming is common, so retention does not imply exclusivity.

Financial Condition & Balance Sheet

What assets are not fully recognized? Internally developed software, models, brand, trained user habits, plugins, templates, customer relationships, and enterprise integrations are largely expensed rather than capitalized. This understates the economic asset base and inflates accounting ROIC relative to a replacement-cost view. It does not eliminate the evidence of strong cash returns.

Off-balance-sheet liabilities? Ordinary purchase commitments, leases, legal contingencies, model/content licensing, and possible regulatory remedies matter. The more material economic commitment is ongoing R&D, cloud, and inference expense required to maintain relevance. It appears in operating expense rather than debt. Unrecognized SBC of $4.21 billion is expected to be expensed over 2.55 years.

How conservative is accounting? The filings show no late reports, restatement, material control weakness, or non-reliance event in the five-year census. Software R&D is expensed, which is conservative for current profit but complicates ROIC. Non-GAAP profit excludes SBC and acquisition-related items, so investors should retain an after-SBC measure. The shift to one segment and reduced ARR granularity is a transparency concern rather than evidence of misstatement.

How capital intensive is it? Physical capex is extremely low: $201 million TTM, under 1% of revenue. Economic reinvestment is high through $4.49 billion of TTM R&D, $6.96 billion of sales and marketing, hosting, data centers, and inference.

Balance-sheet strength? Cash and short-term investments of $5.63 billion against $6.65 billion of debt imply $1.02 billion net debt. Adobe has an undrawn $1.5 billion revolver, no financial covenants on senior notes, and substantial deferred revenue. Liquidity is ample. The shift from $385 million net cash at FY2025 year-end shows that acquisitions and repurchases are consuming flexibility.

Capital Allocation & Management

How much FCF is generated and how is it used? TTM headline FCF is $10.28 billion and after-SBC owner earnings about $8.25 billion. Adobe pays no dividend. It uses cash primarily for repurchases and acquisitions. FY2024 and FY2025 buybacks exceeded headline FCF, while the share count fell materially.

What is the philosophy? Management has favored aggressive repurchases to offset dilution and shrink the share base, with selective acquisitions to widen workflows. H1 FY2026 price discipline improved: $4.59 billion retired 16.6 million shares at an implied $276.39, compared with $432.69 per share in the prior-year period.

Significant acquisitions? Semrush closed for $1.874 billion, 3.9x acquired ARR; 66.8% of consideration is goodwill. Topaz is pending with undisclosed terms. The abandoned Figma transaction cost a $1 billion termination fee and remains evidence of prior willingness to pay a very high strategic price.

Is Adobe issuing large amounts of stock to insiders? SBC was $2.03 billion TTM, 8.1% of revenue. RSUs outstanding rose from 7.9 million at FY2025 year-end to 12.8 million at Q2, partly from grants and Semrush. Buybacks more than offset issuance, but capital used to do so is not a free shareholder return.

Compensation and motivation? FY2025 CEO pay was $51.2 million. FY2026 performance shares depend on relative TSR and total ARR growth, not ROIC or after-SBC FCF. Say-on-pay received only 50.65% support among votes for/against. Insider activity is mixed: director David Ricks bought 10,000 shares at $194.51; the CEO and former CFO sold at lower prices, though stated tax/estate planning and personal circumstances can matter.

Valuation & Market Data

Security form and distributions: ADBE is Nasdaq-listed common equity, not an ADR, MLP, or K-1 issuer. Adobe pays no dividend.

Profitability and cash conversion: TTM net income is $7.23 billion versus CFO of $10.48 billion. The positive gap reflects noncash charges and working capital; no persistent adverse divergence is visible. FCF is $10.28 billion and after-SBC owner earnings $8.25 billion.

Current capitalization: At $263.71 and 397.5 million shares, market cap is approximately $104.8 billion. Adding $1.02 billion net debt gives enterprise value near $105.9 billion. That is about 4.2x revenue, 10.3x headline FCF, 12.8x after-SBC owner earnings, and 10.8x the midpoint of FY2026 non-GAAP EPS guidance. The apparent low multiple requires durable cash; it is not a substitute for organic-growth evidence.

Risks & Downside

What could make the shares decline? High-single-digit FY2027 ARR guidance, weak freemium conversion, paid churn, AI costs growing faster than credit revenue, a poor CEO selection, continued executive departures, integration problems, regulatory remedies, or debt-funded buybacks could lower expected cash growth and the multiple simultaneously.

Catastrophic loss? A major privacy/copyright judgment, security incident, or rapid workflow displacement could cause a severe loss. The installed base, recurring revenue, liquidity, broad products, and low capex make an ordinary-cycle bankruptcy or total loss remote. The realistic downside is a long-duration value trap with cash used to buy shares while the franchise slowly weakens.

Recent News & Events

Q2 results on June 11 showed $6.618 billion revenue, $27.10 billion total ARR including Semrush, AI-first ARR above $500 million, and a raised revenue/non-GAAP EPS guide. The company kept total ARR growth at 10.2%, announced Durn’s departure, and described the freemium/pricing reset. On June 18 Adobe broadened agentic creation and third-party-assistant distribution. On June 25 it agreed to acquire Topaz for undisclosed terms. A July filing documented a retention arrangement for Pentland but no CEO appointment. Figma’s August results showed continued rapid growth and AI-credit use. No later Adobe fundamental print superseded Q2 by August 12.

APPENDIX B — Source Appendix

Adobe Inc. (NASDAQ: ADBE) — Source Appendix

Report date: 2026-08-12 Access date for every source below: 2026-08-12 unless otherwise stated.

Source hierarchy and audit conventions

Public regulatory filings and government or standards records receive the highest weight. Issuer filings, earnings materials, and official announcements are primary sources for what a company reported or announced, while product, market-size, user, and competitor claims remain self-reported. Market data and factor APIs are third-party quantitative sources; news reports are secondary.

“Confirmed” below means the cited source supports the wording; it does not mean that forward-looking management claims have been independently proven.

1. Public primary sources — regulatory, government, and standards records

1. Adobe Inc., Form 10-Q for the quarter ended May 29, 2026. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-06-15. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634326000112/adbe-20260529.htm Supports GAAP financial statements; Semrush acquisition accounting; 397.5M shares outstanding as of 2026-06-11; customer-group revenue; AI inferencing in subscription cost; AI training expense in R&D; and competition, third-party-model, licensing, privacy, IP, compute-cost, and monetization risks.
2. Adobe Inc., Form 8-K (Louise Pentland retention letter). Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-07-17; earliest event 2026-07-14. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634326000120/adbe-20260714.htm Supports a temporary retention/severance arrangement that sunsets after a new CEO starts; it does not announce a CEO successor.
3. SEC company ticker-to-CIK mapping. U.S. Securities and Exchange Commission. Continuously maintained; snapshot accessed 2026-08-12. Regulatory reference dataset. https://www.sec.gov/files/company_tickers.json Supports Adobe CIK 0000796343 and ticker identity.
4. ISO 32000-1:2008 — Document management — Portable document format — Part 1: PDF 1.7. International Organization for Standardization. Published 2008-07. International-standard record. https://www.iso.org/standard/51502.html Historical/structural source confirming PDF 1.7 as an international standard.
5. PDF Specifications. Adobe Inc. Publication date not stated; accessed 2026-08-12. Official technical-history page. https://www.adobe.com/accessibility/pdf.html States that Adobe released control of PDF 1.7 to ISO and ISO 32000-1 was published in July 2008.
6. Adobe / Figma merger inquiry. UK Competition and Markets Authority. Case opened 2023-05-03; provisional findings report published 2023-11-30; case closed 2023-12-19. Government competition case page and report. https://www.gov.uk/cma-cases/adobe-slash-figma-merger-inquiry and https://assets.publishing.service.gov.uk/media/656891e55936bb00133167cf/Provisional_findings_report.pdf Historical source for professional-design entry/expansion barriers, switching/migration/retraining costs, collaboration network effects, interoperability/multihoming mitigants, and case closure after the deal was abandoned.
7. Adobe: consumer protection enforcement case. UK Competition and Markets Authority. Published/opened 2026-03-19; initial evidence-gathering timetable runs through September 2026. Government enforcement-case page. https://www.gov.uk/cma-cases/adobe-consumer-protection-enforcement-case Supports only that an investigation into early-termination-fee terms and presentation is open. The CMA explicitly says no infringement finding has been made.

2. Public primary sources — Adobe financial and investor materials

1. Adobe Reports Record Q2 Results. Adobe Inc. Published 2026-06-11. Issuer earnings release. https://news.adobe.com/news/2026/06/adobe-q2fy26-financial-results

2. Adobe Inc. Q2 2026 Earnings Call — Corrected Transcript. FactSet CallStreet transcript hosted by Adobe Investor Relations; speakers include Shantanu Narayen, David Wadhwani, Anil Chakravarthy, and Steve Day. Event dated 2026-06-11. Company-hosted third-party transcript of issuer call. https://www.adobe.com/cc-shared/assets/investor-relations/pdfs/adbe-q2fy26-transcript.pdf The transcript, release, and 10-Q together confirm $6.618B revenue, reported/constant-currency growth, EPS, customer-group revenue, $22.27B RPO, $27.10B ending ARR including approximately $480M Semrush ARR, AI-first ARR above $500M, Firefly ARR approaching $300M, MAU disclosures, the freemium/line-optimization trade-off, CFO transition, and management’s CEO-search timetable. Transcript statements are management claims, not audited KPIs.

3. ADBE Q2 FY2026 Investor Datasheet. Adobe Inc. Published 2026-06-11. Issuer investor datasheet. https://www.adobe.com/cc-shared/assets/investor-relations/pdfs/11606202/bu45teegrf.pdf

4. Adobe 2024 Investor Meeting presentation. Adobe Inc. Published 2024 (meeting date in document); accessed 2026-08-12. Issuer investor presentation. https://www.adobe.com/content/dam/cc/en/investor-relations/pdfs/amju76er354.pdf Source for management’s $293B 2027 TAM framework. This source is older than 18 months by the report date and the company does not expose enough underlying methodology to validate overlap or convert it into current market share.

5. Adobe Reports Record Revenue in Q1 Fiscal 2022. Adobe Inc. Published 2022-03-22. Issuer earnings release. https://news.adobe.com/news/news-details/2022/adobe-reports-record-revenue-in-q1-fiscal-2022

6. Adobe Reports Record Q4 and Fiscal 2023 Revenue. Adobe Inc. Published 2023-12-13. Issuer earnings release. https://news.adobe.com/news/news-details/2023/adobe-reports-record-q4-and-fiscal-2023-revenue

7. Adobe Reports Record Revenue in Q1 Fiscal 2024. Adobe Inc. Published 2024-03-14. Issuer earnings release. https://news.adobe.com/news/news-details/2024/adobe-reports-record-revenue-in-q1-fiscal-2024

8. Adobe Reports Record Revenue in Q2 Fiscal 2024. Adobe Inc. Published 2024-06-13. Issuer earnings release. https://news.adobe.com/news/news-details/2024/adobe-reports-record-revenue-in-q2-fiscal-2024

9. Adobe Reports Record Q4 and Fiscal 2024 Revenue. Adobe Inc. Published 2024-12-11. Issuer earnings release. https://news.adobe.com/news/2024/12/121124-adobereportsrecordq4andfiscal2024revenue

10. Adobe Q1 FY2025 Financial Results. Adobe Inc. Published 2025-03-12. Issuer earnings release. https://news.adobe.com/news/2025/03/adobe-q1fy25-financial-results

11. Adobe Q1 FY2026 Financial Results. Adobe Inc. Published 2026-03-12. Issuer earnings release. https://news.adobe.com/news/2026/03/adobe-q1fy26-financial-results

12. Adobe Leadership Update. Adobe Inc. Published 2026-03-12. Issuer governance announcement. https://news.adobe.com/news/2026/03/leadership-update

13. Adobe FY2025 Form 10-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-01-15. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634326000003/adbe-20251128.htm

14. Adobe FY2024 Form 10-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2025-01-13. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634325000004/adbe-20241129.htm

15. Adobe FY2023 Form 10-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2024-01-17. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634324000006/adbe-20231201.htm

16. Adobe FY2022 Form 10-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2023-01-17. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634323000007/adbe-20221202.htm Items 13–16 support the FY2022–FY2025 revenue, margin, operating-expense, net-income, cash-flow, capital-expenditure, SBC, share-count, repurchase, debt, and Figma-termination-fee history. FY2022–FY2023 are older than 18 months and are used only to build the historical series.

17. Adobe Q2 FY2026 earnings release, Exhibit 99.1 to Form 8-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-06-11. Furnished issuer earnings exhibit. https://www.sec.gov/Archives/edgar/data/796343/000079634326000109/adbeex991q226.htm

18. Adobe Q1 FY2026 earnings release, Exhibit 99.1 to Form 8-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-03-12. Furnished issuer earnings exhibit. https://www.sec.gov/Archives/edgar/data/796343/000079634326000048/adbeex991q126.htm

19. Adobe Q4 FY2025 earnings release, Exhibit 99.1 to Form 8-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2025-12-10. Furnished issuer earnings exhibit. https://www.sec.gov/Archives/edgar/data/796343/000079634325000135/adbeex991q425.htm Items 17–19 are the authoritative public record for initial/current FY2026 guidance, Q1/Q2 ARR, ARR revaluation, and customer-group guidance.

20. Adobe 2026 proxy statement. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-03-05. Definitive proxy statement. https://www.sec.gov/Archives/edgar/data/796343/000079634326000043/adbe-20260227.htm Supports executive compensation, pay-ratio, performance metrics, equity-plan share increase, ownership, and governance claims.

21. Adobe annual-meeting and repurchase-authorization Form 8-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-04-20; meeting/event 2026-04-15. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634326000101/adbe-20260415.htm Supports vote totals and the additional $25B repurchase authorization.

22. Adobe FY2026 performance-share and cash-incentive-plan Form 8-K. Adobe Inc. / U.S. Securities and Exchange Commission. Filed 2026-01-30; event 2026-01-26. Regulatory filing. https://www.sec.gov/Archives/edgar/data/796343/000079634326000019/adbe-20260126.htm

23. SEC Companyfacts for Adobe Inc., CIK 0000796343. U.S. Securities and Exchange Commission. Continuously maintained; accessed 2026-08-12. Regulatory XBRL dataset. https://data.sec.gov/api/xbrl/companyfacts/CIK0000796343.json

24. Autodesk Q1 FY2027 Form 10-Q. Autodesk, Inc. / U.S. Securities and Exchange Commission. Filed 2026-05-29. Regulatory filing. https://www.sec.gov/Archives/edgar/data/769397/000076939726000044/adsk-20260430.htm

25. ServiceNow Q2 2026 Form 10-Q. ServiceNow, Inc. / U.S. Securities and Exchange Commission. Filed 2026-07-23. Regulatory filing. https://www.sec.gov/Archives/edgar/data/1373715/000137371526000076/now-20260630.htm

26. Intuit Q3 FY2026 Form 10-Q. Intuit Inc. / U.S. Securities and Exchange Commission. Filed 2026-05-20. Regulatory filing. https://www.sec.gov/Archives/edgar/data/896878/000089687826000025/intu-20260430.htm

Items 24–26, together with the latest Salesforce filing and Figma release, support the directional current comp table. Peer enterprise values were mechanically refreshed to August 11 prices using filed spot shares and net debt; differing fiscal calendars and billing changes limit precision.

Items 5–12 are used mainly to date historical price events. The 2022–2024 releases are older than 18 months and are appropriate only for historical context, not the current operating conclusion.

3. Public primary sources — Adobe product and transaction announcements

1. Adobe to Acquire Figma. Adobe Inc. Published 2022-09-15. Issuer M&A announcement. https://news.adobe.com/news/news-details/2022/adobe-to-acquire-figma
2. Adobe Unveils Future of Creative Cloud with Generative AI as a Creative Co-Pilot in Photoshop. Adobe Inc. Published 2023-05-23. Issuer product announcement. https://news.adobe.com/news/news-details/2023/adobe-unveils-future-of-creative-cloud-with-generative-ai-as-a-creative-co-pilot-in-photoshop
3. Adobe and Figma Mutually Agree to Terminate Merger Agreement. Adobe Inc. and Figma. Published 2023-12-18. Joint transaction announcement. https://news.adobe.com/news/news-details/2023/adobe-and-figma-mutually-agree-to-terminate-merger-agreement
4. Adobe to Acquire Topaz Labs. Adobe Inc. Published 2026-06-25. Issuer M&A announcement; terms not disclosed. https://news.adobe.com/news/2026/06/adobe-to-acquire-topaz-labs
5. Adobe Introduces GenStudio for Commerce Media Networks. Adobe Inc. Published 2026-06-17. Issuer product announcement. https://news.adobe.com/news/2026/06/adobe-introduces-genstudio-for-commerce
6. Introducing Adobe Brand Visibility: A Unified Solution for the AI Search Era. Adobe Inc. Published 2026-06-17. Issuer product announcement. https://news.adobe.com/news/2026/06/introducing-adobe-brand-visibility

These sources confirm launches, stated functionality, platform availability, and announced transactions. They do not establish adoption, incremental ARR, retention, or economic returns.

4. Public primary sources — competitors

1. Figma Q2 2026 Press Release. Figma, Inc. Published 2026-08-05. Issuer earnings release containing unaudited quarterly financial statements and operating KPIs. https://investor.figma.com/files/doc_financials/2026/q2/Figma-Q2-26-Press-Release.pdf Supports the cited $370.1M revenue, 48% growth, 136% NDR, customer counts, AI-credit usage, expense, cash, and gross-margin figures. These are issuer-reported; the audit could not fetch the PDF through the web parser but the source URL and local research record were available. A final memo using these figures should prefer Figma’s filed Q2 10-Q/8-K if available.
2. A transformative year for Canva: 2025 in review. Canva Team / Canva. Publication date not exposed on the page; covers 2025 and was available by 2026-08-12. Company newsroom review. https://www.canva.com/newsroom/news/canva-2025-wrap/ Self-reported 260M monthly users, $3.5B revenue, and free Affinity launch; Canva is private and these figures are not filing-reconciled.
3. Welcoming MangoAI and Cavalry as we expand our AI capabilities and professional creative suite. Canva Team / Canva. Published 2026; exact page date not exposed. Company acquisition/product announcement. https://www.canva.com/newsroom/news/mangoai-cavalry-acquisition/ Self-reported acquisitions, more than 5M Affinity downloads, and more than 24B uses of Canva AI tools.
4. Introducing Canva AI 2.0: Reimagining how the world creates. Canva Team / Canva. Published 2026; exact page date not exposed. Company product announcement. https://www.canva.com/newsroom/news/canva-create-2026-ai/ Supports layered/editable output, agentic editing, persistent memory, connectors, scheduling, web research, brand intelligence, Sheets AI, and Canva Code 2.0.
5. More from Canva Create: Canva Offline, Pro Design, Learn Grid & Print Shop. Canva Team / Canva. Published 2026; exact page date not exposed. Company product announcement. https://www.canva.com/en_in/newsroom/news/canva-create-2026-launches/ Supports free Cavalry, Affinity adoption, and professional-workflow integrations. Product scope is self-described and does not prove Adobe churn.

5. Public third-party quantitative and market sources

1. ADBE historical OHLCV and EMA CSV. AZI Trading. Snapshot through 2026-08-11, downloaded 2026-08-12. Third-party market-data CSV. https://azitrading.com/controls/download-data.php?t=ADBE Supports cited price levels, returns, range, EMAs, volume, beta/alpha, and peer-day calculations. Price calculations were independently reproduced from the saved CSV. AZI is not an exchange-primary record.
2. ADBE stock loadings. FactorsToday. Model date 2026-07-31; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/stock-loadings/ADBE
3. ADBE leaderboard. FactorsToday. Snapshot through 2026-08-11; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/leaderboard/ADBE
4. ADBE stock info. FactorsToday. Snapshot through 2026-08-11; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/stock-info/ADBE
5. ADBE stock-specific volatility. FactorsToday. Model date 2026-07-31; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/stock-specific-vol/ADBE
6. ADBE related stocks. FactorsToday. Downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/related-stocks/ADBE
7. Historic factor returns. FactorsToday. Snapshot through 2026-08-11; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/factor-returns/historic
8. Intraday factor returns. FactorsToday. Snapshot dated 2026-08-11; downloaded 2026-08-12. Third-party quantitative API. https://www.factorstoday.com/api/factor-returns/intraday FactorsToday confirms empirical exposures and regime measures only. Loadings are historical model estimates, not business facts. The short-window leaderboard returns are annualized; raw period returns must come from price data. FactorsToday stock-info beta/alpha and AZI beta/alpha are the same computation and are not independent confirmations.
9. Adobe (ADBE) Short Interest & Short Float. MarketBeat. Report date 2026-07-15; page updated in July/August 2026; accessed 2026-08-12. Third-party market-data aggregation. https://www.marketbeat.com/stocks/NASDAQ/ADBE/short-interest/ Supports 20.265M shares short, 5.11% of float, and 3.8 days to cover. Keep the source/date attached; this was not reconciled to an exchange ticker-level file.

6. Public secondary sources

1. “US software stocks rebound, seeking to loosen AI’s grip.” Niket Nishant / Reuters, republished by Investing.com. Published and updated 2026-05-19. News report. https://www.investing.com/news/stock-market-news/us-software-stocks-attempt-a-rebound-as-investors-reassess-ai-risks-4699085 Appropriate for contemporaneous market-driver color, not company fundamentals.

2. “Salesforce Surges 7%, ServiceNow Jumps.” Publisher/author not confirmed in the audit because the page returned HTTP 429. Published 2026-07-27 according to the research-log record. Secondary market commentary. https://finance.yahoo.com/markets/stocks/articles/salesforce-surges-7-servicenow-jumps-183633422.html Replace with directly calculated peer returns or an accessible contemporaneous source in the final memo; do not rely on it for a unique claim.

   The Stock Image decline claim in this note remains unverified and must not appear as fact.

Evidence review: confirmed claims, contradictions, and required qualifications

Confirmed/reconciled

• Core Q2 revenue, EPS, customer-group revenue, RPO, total ARR, Semrush ARR added at close, AI-first ARR, Firefly ARR, MAU, and leadership-transition claims reconcile across Adobe’s release, transcript, datasheet, and/or 10-Q.
• The 10-Q explicitly identifies AI inferencing in subscription cost, AI training in R&D, and third-party-model/compute/legal/monetization risks. This is stronger evidence than paraphrasing product announcements.
• The open-PDF correction is valid: PDF is an ISO standard; Adobe does not legally control a closed PDF format.
• CMA case status/date/timetable are correct, provided the final report states that no infringement finding has been made.
• Price, EMA, return, range, and peer-day figures were reproduced from the saved AZI CSV. FactorsToday values match the saved API responses, with short-horizon annualization properly disclosed.
• Canva’s 260M MAU, $3.5B revenue, 5M-plus Affinity downloads, 24B-plus AI uses, and current product announcements are accurately attributed as company claims. They are not evidence of paid Adobe churn.

Contradiction / tension log

2. Strong reported quarter versus negative price reaction: Q2 revenue, AI-first ARR, and raised revenue/EPS guidance were strong; management simultaneously left total ARR growth at 10.2% including Semrush and accepted a near-term ARR cost from freemium/deferring line optimizations. This is a genuine evidence tension, not an accounting contradiction, and supports explaining why a headline beat did not resolve the organic growth debate.
3. PDF moat wording: The prior baseline’s “file-format captivity” language can survive only as workflow familiarity/interoperability friction. Any suggestion that Adobe owns or controls PDF as a closed standard contradicts Adobe’s own technical history and ISO.
4. Consumer cancellation friction: CMA is investigating whether certain early-termination-fee terms/presentation violate UK consumer law. This creates a tension with interpreting all retention as product captivity, but it is not proof that Adobe infringed the law or that reported retention is contract-driven.
5. CEO succession: Management aimed to have a successor influence FY2027 planning; the July 8-K describes retention protection around a future CEO start. No primary source in the audit names a successor. State only that no successor announcement was located through 2026-08-12, not that none could exist.
6. Private-company metrics: Canva’s scale and adoption figures are self-reported without public filing reconciliation. Figma’s figures are public-company issuer metrics and should be tied to its 10-Q/8-K when possible. Neither set independently proves Adobe share loss.

Unsupported or over-precise claims to remove, relabel, or source better

2. Figma product-extension list (“agents, production code, governance, motion, 3D, and generative plugins”). The financial metrics are supported by the Q2 release; every product in this broad list is not individually evidenced in the log’s cited release. Shorten to features documented in the release or add direct Figma product citations.
3. “Approximately 89% gross-margin heritage.” Plausible from Adobe financials, but the Industry entry provides no period or calculation. State the specific period and show the filing-based calculation, or omit the number.
4. Late-July sector-rally article. The Yahoo source was inaccessible during audit. The independently calculated peer-day return series is sufficient and preferable; drop the Yahoo citation if no accessible replacement is added.
5. Causal event-map language. Price moves are verified; causes are interpretations. The 2021–2022 broad software de-rating and July 2026 AI-risk rotation are the least issuer-specific attributions and should remain explicitly qualified. Earnings-day peer divergence strengthens but does not prove causality.
6. “No new fundamental print superseded Q2.” Reasonable after the public-source sweep but not proven by ROIC’s empty news result. Word as “the primary-source sweep located no later earnings release through 2026-08-12.”
7. “Q2’s switch to one reportable segment.” Date correction required: the Q2 10-Q says Adobe combined its former segments in the first quarter of FY2026, not in Q2. Product-level visibility remains reduced, but the timing should be corrected.
8. “AI-first ARR and Firefly ARR are too small to offset broad seat pressure.” The 1.8% and 1.1% ratios to total ARR are valid calculations, and Firefly is included within AI-first ARR rather than additive. “Too small to offset” and “broad seat pressure” are interpretation; Adobe has not disclosed seat-pressure or churn data. Relabel and avoid adding the two ARR measures.
9. Sequential organic net-new ARR acceleration. Q1 approximately $400M and Q2 approximately $560M are reproducible acquisition-subtraction calculations, but they are not company-disclosed organic KPIs and one sequential comparison is not proof of a durable reacceleration. Label CALCULATION and preserve seasonality/acquisition caveats.
10. FY2024 Figma cash-fee normalization. Adding back the $1B termination fee to operating income and FCF is directionally useful and explicitly non-GAAP, but the cash-flow addback does not model any tax effect. Keep it labeled as a directional normalization rather than an exact recurring FCF figure.
11. Insider census totals and transaction interpretation. The normalized ledger derives from SEC Form 4/4-A filings, but the log cites only local artifacts for the aggregate counts. Preserve the code distinctions (P purchase, S sale, F withholding, M exercise) and add the SEC company-filings index or representative filing URLs if these details enter the final memo. “Bullish” or “negative” readings are interpretation, not filing facts.

Date-integrity check

• Sources older than roughly 18 months are Adobe’s 2022–2024 historical earnings/M&A/product releases, the 2023 CMA Adobe/Figma record, the 2008 ISO standard, and the 2024 investor-meeting TAM deck. The first three groups are appropriately used for historical event/moat context.
• The 2024 TAM deck is the only stale source used for a current market-opportunity figure. Retain only with “management-defined 2027 TAM from the 2024 investor meeting” and the existing methodology/overlap warning; do not present it as a fresh independent TAM.
• Current operating, competitive, market, and regulatory conclusions otherwise use 2025–2026 sources, predominantly Q2 FY2026 materials and August 2026 competitor updates.
• The FY2022 and FY2023 10-Ks are stale only in the date-integrity sense; they are appropriately used for the explicitly historical financial series. Current balance-sheet, guidance, compensation, and governance conclusions use FY2025/Q2 FY2026 filings and 2026 proxy/8-K evidence.

Final evidence standard for the memo

• Every Adobe financial number should cite the Q2 10-Q, with the release/datasheet/transcript used for non-GAAP, ARR, MAU, and forward-looking management KPIs.
• Every management assertion should remain attributed to management. Every calculated organic proxy should show its formula and assumptions.
• Competitor scale/adoption claims should say “Figma reported” or “Canva says.” No competitor adoption statistic should be converted into Adobe churn or market-share loss without cohort/share evidence.
• Interpretations about moat, factor positioning, capital-cycle supply, or market causality must remain labeled as analysis rather than fact.

Disclaimer: This material is for general informational and educational purposes only. It is not investment advice, a recommendation, or an offer to buy or sell any security. Information may be incomplete or become outdated; readers should conduct independent diligence.