The stock market doesn’t lie, even when the earnings do. On March 12, when Shantanu Narayen announced his departure as Adobe’s CEO after 18 years, the company had just reported a revenue beat that would make most enterprises weep: $6.4 billion in quarterly revenue, up 12.1% year over year, crushing analyst expectations by $120 million. Annualized revenue from AI-first products more than tripled. Every metric of success screamed upward. The stock dropped 7.6% in extended trading anyway.
This is the real story, not the resignation itself. The market was saying something that Narayen and the board clearly understood: Adobe has won the earnings game and lost the future.
The Subscription Model That Worked Too Well
Narayen’s tenure represents one of the most successful transformations in software history. When he became CEO in 2007, Adobe was still fundamentally a licensing business, selling expensive one-time purchases of Photoshop and Illustrator to a finite universe of professionals. The subscription model, especially Creative Cloud, solved that problem with ruthless elegance. Customers became recurring revenue. Churn became predictable. Margins became pristine. For fifteen years, it was the operating model that kept Wall Street happy and Adobe shareholders wealthy.
Now it’s the strategic straightjacket that may be slowly suffocating the company.
The Earnings That Looked Like A Death Knell
Here’s what nobody really wants to say out loud: the growth numbers in that earnings report were signs of pathology as much as health. Tripling AI revenue year over year sounds great until you recognize what it actually means. Adobe is spending ferociously to build and acquire AI capabilities. The company is accelerating its integration of generative tools into Firefly, its competing foundation model. It’s doing this because it understands what investors suddenly grasped on March 12: the existing Creative Cloud subscription model is being outflanked.
Midjourney generates sophisticated images in seconds. Runway handles video editing with generative capabilities that professional editors could only dream about two years ago. Open-source models like Stable Diffusion put industrial-grade creative AI in the hands of anyone willing to run it locally. The moat that Adobe spent fifteen years digging around recurring subscriptions is being undermined by a different fundamental shift: the move from software as a scarce resource to software as a commodified capability.
This isn’t disruption in the sense of a startup stealing customers from below. This is a category-level transformation where the value proposition itself becomes questionable. Why should a freelancer pay $65 a month for Photoshop when they can get comparable results from a free AI image generator for prompting and then refine with cheaper tools? Why should a video production company lock itself into Adobe’s subscription treadmill when Runway exists?
The Succession Problem Nobody Mentions
Narayen’s transition to board chairman echoes something he himself mentioned in his memo to employees: a conscious parallel to founders John Warnock and Charles Geschke, who made similar moves. It’s meant to sound like graceful torch-passing, the kind of institutional continuity that reassures stakeholders. But there’s a darker reading available.
Narayen is 62. He has successfully navigated one of tech’s most difficult transitions, from perpetual licensing to subscriptions. He has accumulated enormous personal wealth and influence. What he does not have is a clear roadmap for navigating the next one. The next phase of Adobe’s existence isn’t about executing a model that already works. It’s about fundamentally rethinking what Adobe is when AI makes the existing model optional.
That’s a problem fit for a new CEO to own. And its failures.
The SaaS Category Is Having An Existential Moment
Adobe’s crisis is not unique. It’s the visible manifestation of a question haunting the entire software-as-a-service category: does generative AI enhance legacy software or does it cannibalize it?
This question sits at the intersection of two competing narratives that Wall Street has been trying to hold simultaneously. One says: AI will drive unprecedented software adoption and productivity gains, making SaaS companies more valuable than ever. The other says: foundation models and commodified AI will compress margins and disintermediate software companies by making specialized tools redundant.
Both are true. The problem is they’re true at different timescales and for different customers. Adobe’s enterprise customers might genuinely benefit from integrated Firefly capabilities inside Creative Cloud, making the subscription more valuable. But at the margins, at the long tail of freelancers and small studios and solo creators, the subscription economics become impossible to defend.
The stock market understands this margin collapse is real. That’s why Adobe, like Salesforce, Microsoft, and every other enterprise software company with a visible AI strategy, has watched its valuation expand and contract wildly as investors struggle to price a world where the rules fundamentally change.
Who Benefits From This Transition
The cruel truth is that this isn’t an Adobe problem. It’s a software problem. And it disproportionately benefits the two constituencies that already controlled most of tech capital: foundation model companies and the infrastructure providers running them.
OpenAI, Anthropic, and the other frontier labs don’t have to build the distribution network that took Adobe thirty years to construct. They don’t have to maintain backwards compatibility or deal with millions of paying customers on thousands of configurations. They get to sell capability directly to a world that suddenly believes capability itself is the limiting factor.
Meanwhile, the cloud infrastructure providers, Amazon, Microsoft, Google, get to double-dip. They sell the compute to train and run the models. Then they sell the same models back as services to customers. Then they sell the compute to integrate those services into legacy enterprise software. It’s a three-move checkmate disguised as a productivity boom.
What Comes Next Isn’t Obvious
Narayen’s resignation might have been timed to the earnings beat as a form of institutional face-saving, a way of saying: I’ve left the company in good financial health, and succession is orderly. But it might also be a form of clarity. The next ten years at Adobe won’t look like the last ten. The next CEO will have to make decisions that the current business model doesn’t afford. They’ll have to cannibalize existing revenue streams. They’ll have to compete on something other than distribution lock-in and subscription stickiness. They’ll have to decide whether Adobe’s future is as a creative tool company or as a platform company or as something that doesn’t yet have a clear category.
That’s a different CEO’s problem now. Narayen leaves having won the earnings game comprehensively and clearly understood that winning the earnings game is no longer the same thing as winning the future. The market understood it too. Sometimes the market’s slowness to react masks a sudden clarity underneath. In this case, it was the clarity to recognize that tripling AI revenue while the stock falls nearly 23% this year isn’t the sign of a company that has solved anything. It’s the sign of a company that has finally seen the problem clearly and has no more excuses.
