Adobe at One Billion Users: Can AI Be Sold Separately?

Two years ago the standard line on Adobe was that a free image generator would eat it alive. Type a sentence, get a decent picture, skip the $50 to $80 a month for Creative Cloud. Adobe just reported the quarter that answers whether that fear was right.

Monthly active users crossed one billion in the third quarter, up more than 20% from a year earlier. Revenue hit a record $6.76 billion, up 13%. And the stock still trades at 13.9 times trailing earnings, against a five-year average of 36.2.

My take: the free-tool threat did not disappear, it got absorbed into a much larger, mostly unpaid funnel, and the open question is no longer adoption. It is whether Adobe can keep converting a fraction of a billion free users into paying ones fast enough to justify the multiple it already lost.

A billion users, a familiar discount

Adobe’s non-GAAP earnings per share was $6.13 for the quarter, up 15%, while GAAP net income came in at $1.83 billion, or $4.62 per diluted share, against $1.77 billion and $4.18 a year earlier. Subscription revenue, the core of the business, was $6.58 billion versus $5.79 billion twelve months prior.

That is roughly 13% to 14% growth almost everywhere you look in the income statement, for a company whose critics have spent two years arguing growth was about to break.

It didn’t break this quarter. Full-year revenue guidance moved up to a range of $26.576 billion to $26.626 billion, and non-GAAP earnings-per-share guidance rose to $24.45 to $24.50, both increases from where Adobe started the year.

Investors sold the stock after the report anyway. Adobe trades around $248.92 as I write this, 32.5% below its 52-week high of $369, and the market has priced in a level of doubt the quarter’s actual numbers do not obviously support.

What the freemium bet actually proved

The bear case two years ago was specific: free generative tools would commoditize image and video creation, and nobody would keep paying Adobe hundreds of dollars a year for what a chatbot could approximate for nothing. Instead, Adobe leaned into free.

Creative freemium monthly active users crossed 100 million in the quarter, up more than 70% year over year. That is not a rounding error. It is Adobe deliberately building the exact on-ramp its critics assumed would kill it, betting that scale plus a conversion funnel beats trying to wall off AI features behind a paywall from day one.

AI-first annual recurring revenue grew more than 150% year over year. It is still a small slice of a company doing $23.8 billion in trailing annual revenue, and Adobe has not disclosed the absolute dollar figure. A percentage growth rate off an undisclosed small base tells you direction, not scale, and I’d flag that gap in disclosure before leaning too hard on the 150% number.

I read the freemium numbers as evidence the hollowing-out thesis was too simple, not that it was wrong in spirit. A meaningful share of people who would once have paid for Photoshop or Premiere from day one clearly aren’t anymore. Adobe just found a way to keep them inside its ecosystem instead of losing them to a competitor entirely, and it is now working to charge a smaller number of them more, later.

The mechanics of that conversion matter more than the headline user count. Every one of those hundred million freemium users represents a person Adobe can market Firefly credits, storage upgrades and Acrobat AI features to inside a product they already use daily, at close to zero incremental distribution cost. A standalone AI image startup has to buy that same attention with advertising dollars. Adobe already owns it. That distribution advantage is the part of the old moat that free generators never actually threatened, and it is the part that shows up in AI-first ARR growing faster than the free user base itself.

Free tools didn’t hollow out the suite

The part of the bear case that has not played out is total revenue collapsing under freemium cannibalization. If Adobe were losing its paid base to free substitutes at scale, total company growth would be decelerating toward zero, not holding in the 12% to 13% range it has sat in for several consecutive quarters now.

It’s the steadiness that is the tell here, more than any single quarter’s beat. A company this size, in a category a new technology is supposedly commoditizing, does not usually put up the same growth rate quarter after quarter after quarter. Either the substitution effect is smaller than advertised, or Adobe’s pricing and product moves are offsetting it in roughly equal measure. I can’t fully separate those two explanations from the outside, and that is the one real uncertainty I’d flag in this whole analysis.

I compared this dynamic to what I found looking at the AI duopoly case for AMD: a business people assumed a new technology would flatten instead showing the new technology absorbed into its existing moat. It’s not a perfect parallel, but the mechanism, incumbency plus distribution beating a raw feature advantage, rhymes.

A CEO transition nobody priced in

Shantanu Narayen, Adobe’s CEO since 2007, will move to executive chair on December 1. Anil Chakravarthy, currently president of Adobe’s Customer Experience Orchestration unit, becomes president and CEO on that date, with a stated focus on what Adobe is calling agentic software across its creative, productivity and customer-experience lines.

Leadership transitions at companies this size are rarely clean events for a stock, and the market’s initial reaction after the announcement was negative. I don’t think a single management change explains Adobe’s valuation gap on its own, but it adds a layer of uncertainty on top of the AI-pricing question at exactly the moment investors were already nervous, and that combination is probably why the sell-off after this report was sharper than the numbers alone would justify.

Chakravarthy’s background is on the enterprise and customer-experience side of Adobe’s business, not the creative suite that built the company’s original brand. That is a deliberate signal about where Adobe thinks its next growth phase sits: selling agentic workflow tools to large customer-experience and marketing organizations, not just seat licenses to individual designers. Whether that bet pays off is a multi-year question, and it is fair to say the market has not yet decided how to price a CEO whose most relevant experience is outside the product line Adobe is best known for.

A P/E far below its own history

Here is the number that makes this stock interesting to me: 13.9 times trailing earnings against a five-year average of 36.2. Price to sales tells the same story, 3.9 now versus 10.1 on average over five years. That is one of the largest gaps between a stock’s current multiple and its own history that I’ve looked at this year for a company still growing revenue in the low double digits.

Wall Street’s own ratings reflect the same split personality. Only 38% of the 21 analysts covering Adobe rate it a buy, and the average price target sits at $274, just 10% above where the stock trades today. Compare that to a name like Microsoft, priced closer to its own history, and Adobe looks like the market’s clearest case of a profitable, growing software company being valued as if its moat were already gone.

Metric	Now	Five-year average
Trailing P/E	13.9	36.2
Forward P/E	12.6	—
Price/Sales	3.9	10.1
Analyst buy rating	38% of 21	—
Target price / upside	$274 (10%)	—

That gap is either a mispricing or a warning. A stock this cheap against its own history, with revenue still compounding at 11% and operating margin near 37%, usually means the market has decided the growth won’t last, not that it has mispriced a temporary wobble. I lean toward thinking the market is too pessimistic, largely because the freemium data this quarter argues against the collapse scenario, but I hold that view loosely.

The growth split I would watch next

The specific number I’d want before getting more confident: whether total revenue growth holds in the 12% to 13% band through the next two reports while AI-first ARR keeps compounding well above 100%. If both of those things are true at once, the discount to Adobe’s own five-year multiple looks too wide, and I’d expect at least part of it to close.

The specific downside case: if total growth drops under 10% while AI revenue keeps climbing, that would tell a different story, that AI features are substituting for revenue Adobe used to earn a different way, not adding to it, and the market’s skepticism would be earned rather than overdone. Watch the freemium-to-paid conversion rate specifically, since that is the actual mechanism connecting a billion free users to the earnings-per-share number that ultimately sets the stock price.

Until one of those two paths gets clearer, I’d treat this as a name to own in modest size rather than avoid entirely, funded by the very real chance the multiple re-rates once Chakravarthy’s first full quarter in charge removes one layer of the current uncertainty. Between now and then, the earnings call after the next quarterly report, not the transition date itself, is the event most likely to move that multiple, since it is the first one where Chakravarthy’s own priorities show up in the numbers rather than in a press release.

Analysis and opinion only, not investment advice. Figures come from Adobe’s third-quarter fiscal 2026 results filed on SEC EDGAR and posted to its investor site; user and CEO-transition figures were checked against CNBC’s coverage on September 22, 2026.