Adobe: paid workflows at a price I like

Where I think Adobe can keep earning its money

A finished campaign may require editable files, approved typography, repeatable characters, localization and revisions that survive another round of review. Those tasks make professional software more than a source of first drafts. They also define the work a competitor must replace. A defense of Adobe has to identify the work a competitor would find difficult to reproduce.

More generation can increase the amount of finishing work, but the connection to revenue is conditional. A team may produce a larger campaign with the same seats, buy premium usage or reduce staffing without reducing output. The relevant elasticity is paid spend on the complete process. Additional files are useful adoption evidence without being an additional subscription by themselves.

The competitor has moved beyond an image in a chat box. Canva’s Affinity announcement combines photo editing, vector design and page layout at no charge, with AI features tied to premium Canva access. The announcement leaves parity across Adobe’s creative, video, document, and enterprise workflows unresolved. It does make the price of professional editing a real countercase to the assumption that the surrounding work is protected simply because generation is commoditized.

Adobe’s own architecture permits both its Firefly models and partner models inside parts of its product suite. In June it announced AI assistants in public beta across Photoshop, Premiere, Illustrator, InDesign and Frame.io, with workflows such as reorganizing layers, assembling an initial edit and producing versions of a file. The unified Firefly studio remained a private-beta preview. Those product states matter: a launch establishes availability. Retention and profitable adoption require later evidence. Annual product description, June creative-agent announcement

Adobe’s installed workflows can still win this comparison through better collaboration, reliability, integrations and the cost of changing existing production. Those advantages have to remain useful enough to earn a renewal premium. File familiarity can delay replacement; repeated successful work explains why a customer might willingly pay to stay.

The paid business is still growing

Adobe’s fiscal third quarter ended August 28. Revenue reached $6.760 billion, versus $5.988 billion a year earlier. Customer-group subscription revenue was $4.651 billion for Creative & Marketing Professionals and $1.905 billion for Business Professionals & Consumers, with reported growth of 13% and 16%. The latter group gives the thesis a second source of demand beyond professional creative applications. Q3 financial results

The trailing figures combine FY2025 with the first nine months of FY2026 and subtract the comparable FY2025 period. They preserve a full year of activity. The history also needs context: FY2024 included a $1 billion Figma termination payment. Its disappearance helped subsequent cash-flow and profit comparisons. FY2025 filing, latest quarterly filing

Ending ARR of $27.50 billion grew 11.2% on Adobe’s revalued basis, while $22.16 billion of remaining performance obligations grew about 8%. A4 The filing defines the latter as contracted revenue including deferred and unbilled amounts. Neither is collected cash or a renewal guarantee. Acquired Semrush also contributes to reported growth. These are complementary demand checks whose definitions and acquisition basis must remain attached.

AI adoption still has to become paid demand

Adobe disclosed AI-first ARR above $650 million and growth above 150%. It also reported more than a billion monthly active users across creativity and productivity products. Management’s prepared remarks

The $650 million is a floor. Dividing it by total ARR gives a 2.36% lower-bound comparison. The different measure and incomplete coverage prevent using it as an exact revenue share or as the AI business’s upper bound. Nor does that disclosure isolate how much AI revenue replaces an existing subscription, what serving it costs, or how much profit it creates. The billion users include free activity; they are not a billion paying Creative Cloud customers.

The credit rules clarify how usage can become paid demand. Selected plans include unlimited standard generation, while premium features, including partner models, use credits and can require additional capacity. Adobe can package more expensive work differently from routine generation. A credit allowance still reveals neither the delivery cost nor the margin on its consumption. More usage in a bundled tier and more paid premium usage have different economic effects.

The base keeps consolidated growth below the recent reported pace and adds no AI revenue on top of that growth. Its slightly lower cash margin allows some pressure but is not a bottom-up inference-cost model. Paid upgrades, usage growth and fewer seats must be reconciled before claiming that the AI layer improves either growth or profitability.

I’d watch ARR and obligations before older contracts reach recognized revenue. Currency, price, term length, and acquisitions also affect the measures. Slower commitments need a bridge across those factors. EPS boosted by repurchases can’t resolve the next cohort’s demand.

The best evidence follows an existing team through renewal: comparable paid seats and spend, seats removed, upgrades and metered usage, with acquired and currency effects kept consistent. A replacement AI package changes the old contract; only the net increase adds revenue. That bridge would show whether Adobe captures more of a more productive process or loses part of its earlier seat budget.

Consider a hypothetical team that cuts ten seats to eight at the same price because production becomes easier, then adds a usage package. The new package must first replace the revenue from two lost seats before it creates expansion. The customer may still achieve excellent value. My long requires customers to keep paying and leave Adobe cash after the cost of the new usage.

The cash bridge makes the entry attractive

The latest trailing year produced $10.806 billion of operating cash. Cash capital expenditure of $214 million leaves $10.592 billion of free cash flow. I then charge the $2.071 billion of stock compensation added back in operating cash flow, producing an $8.521 billion cash-after-compensation proxy.

Stock compensation is not an immediate cash payment. This deduction treats employee equity as an economic cost before I ask how much cash can support a smaller share count. This conservative author sensitivity adjusts for employee equity. Adobe does not report it as a metric, and the stock purchases required to offset dilution can differ.

Using the 389.2 million shares disclosed for September 18, the entry price implies approximately $92.684 billion of equity value. The proxy represents a 9.19% yield, or 10.88 times annual cash. I also cross-check earnings: the entry is about 13.1 times the midpoint of FY2026 GAAP EPS guidance, $18.145. Share count and cash statements, EPS guidance

The entry appeals because a large existing cash base is available at a modest multiple after an economic compensation charge. That can compensate for uncertainty if paid demand is durable. It cannot make a structurally declining stream safe simply because the current yield is high. Competing workflows can change the price customers pay, so the product comparison belongs in the valuation.

Fewer shares need a cash budget

Repurchases strengthen this case only when they create affordable net retirement. In the first nine months, balance-sheet shares fell from 413 million to 391 million, a 5.33% reduction, while Adobe spent $6.821 billion on repurchases. That is evidence of a shrinking denominator. It is not another return to add to a target price already calculated with fewer shares.

My base case asks for 4% net retirement in each model year. At assumed average repurchase prices of $285 and $330, that requires $4.437 billion and $4.932 billion of net-purchase funding. Rising prices make later retirement more expensive; the model recognizes that.

The first-year budget includes $340 million for the announced Topaz transaction and $1 billion of additional capital flexibility; the second also includes the $1 billion allowance. I set these budgets for the Adobe forecast; they are my assumptions. Unspent cash is left out of the terminal price, so it cannot quietly become an extra valuation benefit.

Acquisitions and repurchases use the same surplus that supports new products and changing delivery costs. An acquisition must add paid demand and contribution; a repurchase must increase each remaining share’s claim at a worthwhile price. The capital budget creates flexibility without assigning either choice a return in advance. If usage economics require more investment, the net-retirement plan must adjust.

How I get to $365

My base case grows the $25.970 billion revenue anchor by 8% annually, reaching a $30.291 billion annual run rate near the evaluation date. Cash after compensation equals 31% of sales, below the observed trailing margin of roughly 32.8%. That produces $9.390 billion of annual cash at a lower cash margin than the trailing year.

With 358.687 million ending shares, the result is $26.18 per share. Fourteen times that cash produces $366.52 in the model, which I register as a rounded $365 central forecast. A 27% net-margin cross-check produces about $22.80 of earnings per share; the exact model price is roughly 16.1 times those earnings. The cross-check does not get added to the cash valuation.

At the entry cash multiple, the base produces approximately $284.76, a 19.6% gain. The central forecast requires a move from 10.88 to 14 times cash as well. Paid growth that survives cheaper workflows would justify part of that revaluation. A high count of free users, generated files or products labeled AI would leave paid durability unresolved.

The reverse calculation shows the condition clearly. With the base share count and a 14-times multiple, Adobe needs about $7.383 billion of annual cash to support a 10% annualized price return from the entry. That is 13.4% below the observed proxy. This room exists because the calculation assumes both net retirement and a higher valuation multiple.

Bear, base, and bull cases

The bear combines fewer paid workflows, lower cash margin and no net retirement, yielding roughly $140. It can occur even if Adobe remains a good tool: customers may obtain adequate results elsewhere or need fewer subscriptions. A large current cash flow makes the transition financeable; it does not eliminate the loss to an investor who pays for durability that fails.

The roughly $530 bull requires teams to keep Adobe for valuable production, buy enough additional capability to expand spend and leave cash after the usage is served. Faster retirement must remain affordable as prices rise. This is a case in which workflow ownership and model aggregation earn contribution, with productivity gains translated into paid demand.

Holding base operating assumptions constant, 11 times cash gives approximately $288, while 17 times gives $445. This spread tells me that recognition by future buyers matters almost as much as modestly changing the growth forecast.

What I will judge before the horizon

I want the latest available FY2028 results to show an annual revenue run rate above $30 billion, cash after the same compensation charge above $9 billion, and comparable ARR growth of at least 7%. I also expect net retirement near the base case’s 7.84% over two model years, achieved without relying on expanding debt. These are my prospective tests.

Two successive reports below 5% ARR growth, GAAP operating margin below 30% or trailing compensation-charged cash below $7.5 billion would materially weaken the dated long. A growing free audience beside those outcomes would strengthen the substitution case. Acquisitions that consume the capital budget without improving paid demand would require the share forecast to change too.

Quarterly renewals and cash will determine whether Adobe keeps being paid for a process that customers can increasingly reassemble elsewhere. The scheduled December 1 transition to Anil Chakravarthy remains a dated strategic checkpoint, with no automatic value assigned to leadership change. A6 I would own Adobe when the complete workflow earns its premium after the customer has a credible alternative. That is a stronger defense than insisting that professional work will never stop needing software.

My long thesis and dated forecast

I favor Adobe when paid growth survives credible cheaper alternatives and recurring cash funds the share reduction. I keep the $365 October 12, 2028 forecast against SPY. The model needs a higher cash multiple as well as operating progress. Renewal spending and cash after usage costs are the evidence that can earn that revaluation.

Assumptions

• USD millions and millions of shares unless stated; prices are USD per original share.
• TTM = annual result + current comparable interim − prior interim. Forward cash margins include economic compensation and company-specific recurring reinvestment.
• Net share reduction is cash-budget constrained. Retained/protected cash has no terminal bonus; gross buybacks are not added to price returns.