Adobe FY 3Q26: A Year of Investments in Strategic Client Acquisition

And the payoff is on the New Management to Deliver!

Please follow our analysis of Adobe FY 3Q26 and the delicate moment the company is in, with a clear strategic bet in mind, but still need to prove itself under the new management.

1. The tension underneath the call:

Revenue and ARR are growing double digits, but net new ARR was lower than usual and RPO growth fell to single digits for the first time since early FY23.

Management’s answer to both is the same — it’s deliberate, we’re stuffing the freemium funnel and we deferred price increases. That’s a coherent story. It’s also unfalsifiable for another two to four quarters.

*We try not to take strong conclusions from Net New ARR down 39%, since last quarter it was up 79%. So definitely was a low print, but also is a volatile number and the company acquisition strategy help to explain short term headwinds.

*We previously though RPO would be more immune to freemium, but as for the call that doesn’t seen to be the case?

2. The reported numbers were actually good

Metric	Q3 FY26	Growth
Total revenue	$6.76B	+13% reported / +12% cc
GAAP EPS	$4.62	+11%
Non-GAAP EPS	$6.13	+15%
Customer-group subscription revenue	$6.56B	+14% / +13% cc
Cash from operations	$2.52B (Q3 record)	—
Shares repurchased	~9.5M	~2,4% of Shares Outs
Business Professionals & Consumers (BP&C)	$1.91B	+16% / +15% cc
Creative & Marketing Professionals (CMP)	$4.65B	+13% / +12% cc

• The $6.76 billion revenue came +1,8% above our estimates.
• We can see clearly that the Freemium AI motion is also driving higher consumption, being subsidized by Gross Margins.
• Expenses also came in higher than expected, growing 14,8% YoY %, with higher R&D than we expected, which indicates company is not standing still in product development!
• Diluted EPS grew 10,7%, driven by 8,3% Operating Income Growth and stock buybacks.
• Important Note: Buyback capacity is enormous and freshly reloaded. The March 2024 authorization is now fully consumed; ~$24.55B remains under an April 2026 authorization — roughly 92% of a full year’s revenue.

Gross Margin -0,6pp YoY.

Operating Margin -1,5pp YoY.

Net Income Margin -2,6pp YoY.

Free Cash Flow was up +15% with +0,6pp better margin than a year ago.

3. The AI disclosure set

This is where Adobe is choosing what to show, and the pattern is informative.

Metric	Disclosed
AI-first ending ARR	>$650M, +>150% YoY
Total MAU (all businesses)	>1 billion, +>20% YoY
BP&C MAU (Acrobat + Express)	>900M, +>25% YoY
Creative freemium MAU	>100M, +>70% YoY
Acrobat AI Assistant MAU	doubled QoQ
Firefly ending ARR (app + credit packs)	+40% QoQ
Brand visibility paid customers	doubled QoQ
CX Enterprise Coworker	1,700+ customers and early adopters
AEM + agentic web apps / GenStudio / AEP+apps	each >20% ARR growth YoY

• AI-first ARR of $650M is ~2.4% of the $27.5B base. Growing 150%+ is genuinely fast, but even a second doubling adds ~$650M — under one full point of total company growth. The AI revenue line is not yet load-bearing. Narayen implicitly conceded this by pivoting to “AI-influenced revenue” — “if you look at the AI influence revenue and what’s being driven by Creative Cloud, all of the core desktop applications, the credit consumption there also is very robust.” That’s the right argument (AI is defending the base, not just adding a new line).
• The 1 billion MAU number is a rollup, and the arithmetic is suspiciously clean. 900M BP&C + 100M creative freemium = exactly 1B. Treat “1 billion monthly active users” as a sum of segment counts, not a deduplicated user count.
• Percentages without denominators. Firefly ARR “+40% QoQ,” Acrobat AI Assistant MAU “doubled QoQ,” brand visibility paid customers “doubled QoQ” — none have a stated base. Doubling from a small number is easy. This is a deliberate disclosure choice and it recurs across every AI metric on the call except AI-first ARR. When they start giving absolute dollars, that’s the signal the base has become defensible.
• The QoQ framing is itself a tell. Healthy, large businesses get disclosed year-over-year. Everything young here is quarter-over-quarter.

4. Guidance

Day characterized the raise as ~$50M at the midpoint versus the June guide, across CMP subs, BP&C subs and total.

• A $50M raise on a $26.6B base is 0.2%. It is a raise, and Day was right to say “we take our guide seriously.” But it is the smallest possible gesture that still counts as one.
• Q4 CMP implies +0.6% sequential growth ($4.65B → $4.68B at the midpoint). BP&C implies ~+1.6%. Total implies ~+1.0% QoQ. For the segment carrying 69% of revenue, essentially flat sequential in the seasonally strongest quarter is the weakest number in the guide — which is exactly why JPMorgan led with it.
• Q4 margin steps down. ~44% non-GAAP in Q4 against ~45% for the year implies Q4 carries above-trend spend — MAX in Miami Beach, AI compute against accelerating credit consumption, and the Topaz Labs close.
• EPS still steps up ($6.13 → $6.30–6.35) on volume and buyback, but the margin line is where the AI cost is showing.
• The ARR guide requires a very large Q4. Working from the company’s own figures: 10.2% on a $25.66B beginning book implies an ending book of ~$28.28B. Ending ARR at Q3 was $27.5B. That leaves ~$780M of net new ARR to be added in Q4 alone, against ~$1.84B added across the first three quarters combined — i.e. Q4 must deliver roughly 30% of full-year net new ARR. Management’s defense is that Q4 is structurally the enterprise quarter and that they pre-signaled this shape on the Q2 call. Both are true. But it is the single largest execution dependency in the guide, and it lands in the same quarter the CEO changes. (Caveat: “ending ARR book of business” and “ending ARR” may not be strictly identical definitions; Day used both terms without reconciling them.)
• The FY raise was smaller than the Q3 beat. Citi caught this. Day’s answer: “nothing changing in the business at all… there is a slight FX headwind that we’re having coming into Q4, and so it’s just netting that off.”

5. The leadership transition

• Anil Chakravarthy becomes President & CEO and joins the Board effective December 1. He joined Adobe in January 2020; by end-2020 Narayen had handed him enterprise field operations across the whole company — a detail Chakravarthy volunteered specifically to rebut the “he’s the enterprise guy” framing.
• Narayen becomes Executive Chair and says he’s committed to a close transition.
• David Wadhwani was thanked “for over a decade of partnership, leadership and passion for our creative community” — phrasing that reads as a departure, and Narayen confirmed the shortlist structure: “We had 2 great internal candidates.”
• Steve Day remains Interim CFO. No permanent CFO was named or asked about.

Read:

• Three senior seats in motion at once — CEO, the creative-business leader, and a CFO role still marked interim — during the quarter that has to carry 30% of full-year net new ARR. Nobody on the call asked about the CFO search, which is a gap in the Q&A rather than in the company.
• Narayen pre-empted the Barclays question before it was asked of Chakravarthy, and the rebuttal he chose is telling: “as it relates to how software will be sold, the individuals and making sure that we delight individuals is going to be even more important moving forward.” The Board appointed the enterprise executive and then had the outgoing CEO stress that individuals matter more than ever. That is a deliberate signal that the freemium/consumer strategy is not being unwound by the successor.
• Chakravarthy’s own stated priorities are notably tactical: close out Q4, host MAX, spend time with community/customers/employees, and “building the leadership team and putting all of that in place.” That last phrase is the one to carry forward — expect further org changes before the new year. He offered no strategic reset, which is the correct posture pre-start date but leaves the strategy question genuinely open until MAX or the Q4 call.

6. Strategy as management framed it

The organizing claim: every SaaS company must transform to “agentic software” the way software went from on-prem to SaaS, and Adobe intends to lead that transition in creativity, productivity and customer experience. Chakravarthy defined what customers mean by agentic software in four parts: (1) interface of the user’s choice, conversational or traditional; (2) model flexibility with intelligent routing to the best model per task; (3) grounding in the customer’s own enterprise data and context to eliminate hallucination; (4) help realizing value, delivered via forward-deployed engineering and value-realization teams.

BP&C / productivity. Acrobat is being repositioned from “PDF tool” to “document productivity platform,” anchored on 400 billion PDFs opened annually and three decades of format expertise. New: the Adobe Productivity Agent; documents converted into interactive reports, summary slides and audio/podcast formats; Knowledge Base and Analyzer for querying across thousands of documents with enterprise governance; Student Spaces globally for higher education (flashcards, quizzes, study guides). Acrobat capabilities are being extended into ChatGPT, Chrome, Claude, Microsoft Edge and WhatsApp. Wins: Amazon, Deutsche Post DHL, Honeywell, JISC, Renault, Los Alamos National Laboratory.

Creative. Firefly as “all-in-one creative AI studio,” now with native music, speech and sound effects generation (commercially safe audio); Create Storyboard and Create Brand Kit for social creators and solopreneurs. Adobe Creative Agent extended into Photoshop and Premiere. New AI-assisted Photoshop editor in beta with a natural-language prompt bar and third-party models.

Topaz Labs acquisition announced (>1M users, Emmy-winning enhancement models, expected to close Q4 subject to regulatory approval). Firefly Enterprise expanded with Firefly Graph EE, Creative Production EE, and a Simulate feature in Brand Intelligence that predicts campaign performance pre-launch. Disney Imagineering is integrating Firefly Foundry. Saudi partnership with the Ministry of Communications and IT and HUMAIN to reach 27 million citizens and residents, plus a culture-tuned image model. MAX in November, Miami Beach, 10,000+ in person. Topaz is state-of-the-art AI enhancement across imaging and video; customers were already using Adobe plus Topaz together and gave strong feedback; natural adjacency to Creative Cloud, Firefly and Firefly Enterprise; closing expected in Q4; more at MAX.

CX Enterprise. Positioned as “the agentic system of record for Customer Experience Orchestration,” three pillars: AEP + native apps for engagement, GenStudio for content supply chain, AEM + agentic web apps for brand visibility. Over 20,000 global enterprises; over 1 trillion experiences delivered annually. Adobe Brand Visibility combines Adobe LLM Optimizer with Semrush AI optimization over a database of ~300 million real-world AI search prompts, tracking brand presence across ChatGPT, Google AI Mode, Copilot and Perplexity.

CX Enterprise Coworker GA’d in June with 1,700+ customers and early adopters. GenStudio for Commerce Media Networks launched at Cannes Lions. Analyst recognition: top leader in two Forrester Waves and two IDC MarketScapes. Wins: Academy Sports, Alpine Racing, BNP Paribas, Humana, IKEA, Jet2, Marriott, MSC Cruises, Publicis, Royal Bank of Canada, Vanguard, Wells Fargo.

Reads:

• Adobe has decided not to fight for the interface layer. Shipping Acrobat into ChatGPT, Claude, Copilot, Edge and WhatsApp, plus routing to third-party models inside Photoshop, is a clear strategic posture: be the workflow, data and trust layer, not the model and not the chrome. The bull case is durability — model price deflation becomes a gross-margin tailwind and Adobe stays relevant regardless of which assistant wins. The bear case is that the UI was the moat; renting distribution inside a rival’s interface means the rival owns the customer relationship and can, at any point, build one layer down. This is the most consequential unhedged bet in the whole call and no analyst pressed on it.
• “Generative engine optimization” is the genuinely new product category here and it’s growing off a tiny base (paid customers merely doubled). The 300M-prompt database is the defensible asset, not the optimizer software.
• The enterprise CX story now has three separate “>20% ARR growth” lines (AEM+agentic web, GenStudio, AEP+apps) — a disclosure structure that lets Adobe report three strong growth rates without revealing the mix or how much is Semrush-acquired.

7. Two Highlights from Q&A

Brad Zelnick (Deutsche Bank) — what’s driving Firefly credit consumption acceleration? Broad usage or specific generation types?
Chakravarthy: the funnel working as designed — acquire via freemium (100M MAU, +70%), increase intensity of AI usage and engagement, and that engagement translates into ARR (+40% QoQ for Firefly app + credit packs). Narayen added specificity: video in particular is seeing heavy usage, and credit consumption inside the core Creative Cloud desktop applications is “very robust.”

The read: Narayen’s addendum is the more valuable half. Video is the highest-credit-consumption modality per generation, so a video-led mix shift mechanically inflates credit consumption without necessarily implying proportional user-count or willingness-to-pay growth — and it raises COGS. Separately, “credit consumption in the core desktop apps is robust” is the strongest available evidence that AI is defending the $18B installed base rather than just building a new one. That’s the argument that matters most for the terminal value, and it’s the one they can’t yet quantify.

2. Ivan Radojicic for Alex Zukin (Wolfe) — last quarter you flagged headwinds from doubling down on freemium and from deferring pricing initiatives. How much came from each? And when do the deferred Creative Cloud pricing optimizations resume?
Narayen answered: the two priorities were singular company focus and impatience to drive MAU; segmentation gives Adobe room to deliver more value within Creative Cloud over time; and — “I’m actually really happy that we didn’t focus on the pricing actions because that, while it may have provided some short-term relief, would not be as critical as continuing to drive new user adoption.”

The read: Both questions went unanswered. No split between the two headwinds, and no timeline for resuming price increases. The phrase “short-term relief” is an unusually candid admission that the pricing lever was available and was consciously left unpulled — which means reported ARR growth currently contains zero price contribution. That cuts both ways: it makes the 11.2% ARR growth more organic than it looks, and it leaves a real, undated, un-quantified lever sitting on the table. The absence of a timeline is the honest read that they don’t yet know when the funnel will be full enough.

8. What management did not answer

Worth keeping as a list for next quarter:

1. Why RPO growth halved to single digits — the freemium explanation doesn’t reach contracted backlog.
2. The split between the freemium headwind and the deferred-pricing headwind (Wolfe asked directly).
3. When Creative Cloud pricing optimizations resume — no timeline, not even a framework.
4. Semrush’s quantified contribution — now permanently folded into blended CX growth lines.
5. Which AI pricing model becomes primary — no commitment.
6. Any named strategic, technical, competitive or financial challenge — asked explicitly by Griffin, answered with an asset inventory.
7. Why Q4 CMP revenue is only ~0.6% sequential — answered with ARR and product commentary instead.
8. Topaz purchase price or financial impact — not disclosed.
9. The permanent CFO search — not raised by anyone.

9. The bull and bear cases, strictly from this transcript

Bull. Revenue +12% cc with a raise, record Q3 operating cash flow of $2.52B, ~45% non-GAAP operating margin, and a $24.55B buyback authorization. AI-first ARR compounding at >150% off $650M. Creative freemium MAU +70% to 100M, feeding a funnel that has a decade-proven conversion machine behind it and hasn’t been monetized yet.

Credit consumption rising inside the legacy desktop base — evidence AI defends the installed base, not just adds a new SKU. Pricing power fully intact and deliberately unused (sure?). Enterprise CX growing >20% across three product lines with a credible new category (generative engine optimization) and a differentiated data asset. An internal, five-year CEO with cross-portfolio ownership and a committed Executive Chair alongside him.

Bear. Net new ARR is down 36–37% and RPO growth has halved, and only one of those has a coherent explanation. Total ARR growth of 11.2% is below revenue growth of 12–13%, which means revenue growth mathematically converges down toward ARR unless net new inflects.

The ARR guide requires Q4 to deliver ~30% of the full year’s net new — in the quarter the CEO changes, with a creative-business leader departing and an interim CFO. AI-first ARR is 2.4% of the base and every other AI metric is disclosed as an undenominated QoQ percentage.

Monetization architecture is explicitly undecided three years in. Semrush disclosure has just been retired. The company is placing its distribution inside ChatGPT, Claude, Copilot and Edge — hedging the interface war by renting shelf space in competitors’ stores. And the only definite answer on when pricing power gets used was that it isn’t being used.

10. What to check next quarter

• Net new ARR in Q4 — did it hit the ~$780M the 10.2% guide implies, and is the mix enterprise or funnel conversion?
• RPO / cRPO — does the Q4 seasonal step-up restore double-digit YoY growth, or does the spread persist? If cRPO keeps outgrowing RPO, ask about contract duration directly.
• AI-first ARR — the $650M base and whether the >150% rate survives a larger denominator.
• First absolute dollar figure for Firefly ARR — the moment QoQ percentages become dollars is the moment the base is defensible.
• Any resumption of Creative Cloud pricing, and whether it arrives as paywall tuning or as a list-price event.
• CMP sequential revenue — whether the Q4 flatness was FX and timing, or the beginning of a trend.
• Leadership: permanent CFO, creative-business leadership, and whatever Chakravarthy’s “building the leadership team” produces.
• MAX in November, which is now carrying the entire forward strategic narrative.

This is an analytical piece only, we tried to let both bear and bull views open.

For Opinions, Valuation and Stock Expected Return, follow us!

Thank you!

CoreValue Research Team