For four years straight, the B2B playbook has been the same. Raise prices every 12 to 18 months. Add an AI SKU you have to pay extra for (another price increases, really). Customers grumbled and paid, because where else were they going to go. They were locked in.
It finally stopped working for Adobe, at least at low end of its market, the prosumer side.
On its Q2 FY2026 call, the company told Wall Street it is deferring ~$500m ARR of Creative Cloud price increases it had already planned for the back half of the year.
Adobe historically has been one of the most pricing-powerful companies in all of software. It raised prices every year for the past 4+ years, with big raises in 2022-2025:
When a business like that voluntarily walks away from a key planned price hike, it’s big deal.
Adobe stock crashed — even though the quarter itself was strong
Adobe posted record revenue of $6.62 billion, up 13% year over year as reported and 11% in constant currency. Non-GAAP EPS came in at $5.96, up 18%. Total ARR hit $27.1 billion, growing 12.5%. The company raised its full-year revenue and EPS guidance, helped by the April acquisition of Semrush, which added $480 million in ARR.
The AI traction is real too, even if it’s hard to distinguish from just plain … revenue:
- Creative freemium monthly active users grew from 50 million to 90 million in a single year
- Firefly ARR grew roughly 50% quarter over quarter and is approaching $300 million
- AI-first ARR across the company grew 3x year over year to over $500 million
- Acrobat AI Assistant paid MAUs grew more than 150% year over year
- Traffic to adobe.com grew more than 40% year over year
So this is not a company in deep trouble on the surface, even if AI is a threat on many vectors.
And yet the stock had already fallen about 38% year to date heading into the quarter, and Citi cut its target to $228 from $264. The market is not reacting to the quarter. It is reacting to what the price deferral reveals.
The official story, and the real one
Adobe’s spin and framing is clean. CEO Shantanu Narayen says the company is going aggressively after a freemium funnel, trading near-term ARR from individual subscribers for a much larger base of free users it can monetize later. The Adobe Reader playbook, applied to Firefly and Express. Get hundreds of millions of people in the door for free, convert them over time.
CFO commentary put real math behind it. Roughly half the cut to second-half ARR comes from deferring Creative Cloud price optimizations. The other half comes from routing more users into freemium experiences instead of direct-to-pay flows. Call it a $500 million implied trim to fiscal 2026 organic ARR.
That story is true. It is also incomplete.
The freemium pivot and the price deferral are the same decision. You cannot raise prices and run a land grab for free users at the same time. The two work against each other. Adobe picked the land grab. The interesting question is what forced that choice, and the answer is sitting in the competitive data.
Adobe has two pricing markets, and they are splitting apart
Adobe’s pricing power has not completely collapsed. It has bifurcated.
At the professional and enterprise tier, the moat holds. Designers live in .AI, .PSD, and .INDD files wired into client review and production pipelines. There is no real substitute for Illustrator on serious vector work or Premiere on broadcast video. Adobe’s posture on 2026 enterprise renewals is to capture the full 2025 list increase, and it is still pushing 8 to 12% on multi-year ETLA deals. No deferral there. The enterprise wins named on the call: SAP, ServiceNow, Workday, Coca-Cola, Merck, Accenture.
At the consumer, prosumer, and SMB tier, the moat is weaker. And that is exactly the tier where the price increase got deferred. The substitutes are now credible and they are a fraction of the price:
- Canva crossed $4 billion in ARR, growing more than 30% a year
- Figma, the deal regulators blocked Adobe from buying, is at $1.2 billion in ARR growing 40%
- Canva made the Affinity suite free. The Canva Pro plus free Affinity bundle now does the job of Creative Cloud for about $15 a month, against Creative Cloud at $55 and up
- AI-native tools were built around generation from day one, not retrofitted
William Blair already downgraded Adobe to Market Perform in March on precisely this thesis. So when Adobe declines to raise prices on the individual tier, it is not purely funnel strategy. Raising prices into a market where Canva-plus-free-Affinity does a lot of the job for a quarter of the cost would simply accelerate the defection. The freemium reframe is the smart way to tell that story. But the constraint is real.
Is this defense or offense? Both.
The analysts who like Adobe here argue the moat is intact where it pays the bills. Switching costs for working professionals are enormous. Proprietary formats are wired into every production pipeline on earth. And Adobe has one thing the AI-natives cannot easily copy: copyright indemnification and commercially safe training data through Content Credentials, which is exactly what large brands and their legal teams require before they will ship generated work. Adobe now trades around 14x forward earnings, the lowest multiple in a decade, which tells you the market has already priced in heavy disruption. If the freemium funnel converts the way Reader once did, the deferral looks like smart offense in hindsight and the stock is cheap.
The bear case is real as well. Adobe is leaning harder on third-party models, including the new NVIDIA Firefly Foundry partnership and support for multiple outside media models inside its own apps. That pushes Adobe toward being an integrator of other companies’ AI rather than the source of it. When Sora, Veo, Midjourney, and Runway own the ideation step, Adobe risks becoming the place you finish work, not where you start it. That is a worse margin position and a weaker pricing position at the same time.
Both can be true. The deferral is defense forced by the low-end threat, and it is offense if the funnel pays off. What it is not is business as usual, and the price page is where that shows up first.
It’s not just Adobe. Everyone raised prices in 2024, 2025 and if they could … 2026.
The reason the deferral stands out is that it broke from a near-universal move. Across the public software leaders, the 2025 and 2026 story was the same lever pulled over and over:
- Microsoft went up on nearly every front: Power BI Pro up 40%, Teams Phone up 25%, Dynamics 365 Business Central up 10 to 15%, a new 5% surcharge just for paying monthly, and a July 2026 round taking Microsoft 365 E3 up about 8% with some Frontline plans up to 43%
- Google Workspace rose 17 to 22% in early 2025, with Gemini bundled in as the justification
- Adobe restructured Creative Cloud into Standard and Pro tiers for effective increases up to 27%
- Atlassian pushed Data Center prices up 15 to 40%, herding customers toward cloud
- ServiceNow runs a mandatory 5 to 10% annual uplift, then layers Now Assist AI on top at a 30 to 45% premium
- Salesforce added 6% on Enterprise and Unlimited in August, on top of the 9% it took in 2023
- At the far end, Broadcom drove VMware bills up anywhere from 150% to over 1,000% through forced subscription migration
The mechanism is identical every time. The AI feature is the justification, the bundle is the delivery, and the customer mostly cannot opt out. All of it landed in the same window that wiped close to $2 trillion off enterprise software market caps in Q1 2026, with investors finally asking the question vendors have no good answer to: if an agent can do the work, why am I still paying per seat.
Two leaders didn’t completely raise prices (although both did in part, where they could):
- Adobe deferred its 2026 increase on its prosumer customers.
- HubSpot went the other direction entirely and cut its Starter tier to $15 a seat to take the low end rather than squeeze it.
Both are betting that in an AI market, owning the customer beats charging them more. Everyone else is still pulling the lever, and the market is repricing them for it.
There will be more. The Golden Age of Endless Annual Price Increases in business software may be coming to an end, partially at least.
Adobe is not an outlier. It is an early, visible example of something happening across the entire B2B + AI landscape at once. The price-increase machine that defined 2022 through 2025 ran on a simple condition: lock-in plus no substitute. AI just broke the second half of that equation in category after category. When a credible AI-native substitute appears, your ability to push price at the exposed tier evaporates, even when it holds at the top of your customer base.
Look at who repriced in just the first months of 2026:
- HubSpot moved Breeze AI agents to outcome-based pricing and cut the price from $1.00 to $0.50
- Anthropic lowered enterprise seat prices for Claude while leaning harder into usage
- SAP announced a shift toward AI consumption pricing
- GitHub Copilot layered token-based AI Credits on top of its per-seat plans
- Clay split its model into platform value and token cost as separate buckets
The companies at the bleeding edge of pricing change in 2026 are not the startups. They are the largest incumbents, because they have the most legacy seat revenue to defend. And there is an ugly margin fact underneath all of it: AI gross margins run closer to 50%, well below the 70 to 80% that classic B2B trained everyone to expect. AI spend is largely cannibalizing existing software budgets rather than adding to them. So you have new substitutes pressuring price at the same moment the new revenue carries worse margins. That is the squeeze Adobe is managing in public.
The risk hiding in a lot of B2B growth rates
Over the last three years, how much of your growth was new customers, and how much was just charging the base more?
For a lot of companies the honest answer is uncomfortable. At Salesforce, price increases now account for up to 72% of forward ARR growth. And Salesforce is not alone in leaning on the lever. SaaS pricing rose roughly 11.4% in 2025, close to five times the 2.7% inflation rate across G7 economies, while corporate IT budgets grew about 2.8%. The gap between what vendors charged and what customers could actually absorb has been filled by the price lever, year after year.
That works right up until it does not. If most of your growth is price, then your growth is only as durable as your pricing power. The day a credible AI-native substitute lands at your exposed tier, the lever jams, and the demand underneath was never as strong as the top line implied. You do not get a slow decline. You get a step change, because the growth was manufactured, not earned.
Adobe could afford to defer because its enterprise moat is real and it has a genuine freemium funnel to fall back on. Most companies running 50, 60, 70% of growth off price increases do not have either. They have a thinning moat and a base that is one good alternative away from pushing back hard. For them, “we can’t raise prices this year” is not a strategy choice. It is the moment the music stops.
Stop assuming your annual price increase is automatic
The lesson from Adobe is not “stop raising prices.” Pricing power is alive and well where you have genuine lock-in and no substitute. Adobe proved that this quarter on the enterprise side.
The lesson is to know which of your tiers still has pricing power and which one just lost it, before the renewal data makes the call for you.
A few things to pressure-test in your own business:
- Map your substitute risk by segment, not in aggregate. Adobe’s enterprise base is fine. Its consumer base is exposed. A blended churn number would have hidden that completely. Run the analysis tier by tier and use case by use case.
- Stop assuming the annual price increase is automatic. For four years it was a free lever. In any category where an AI-native tool now does 80% of the job at 20% of the price, that lever is gone, and pulling it just speeds up the loss.
- If your moat is eroding at the bottom, get there first with free. Adobe’s freemium pivot is defensive in origin but it is the right move. Better to commoditize your own entry tier and own the funnel than to let Canva own it. The mistake is waiting until the data forces your hand.
- Reprice around the value you deliver, not the seats you sell. Gartner expects at least 40% of enterprise software spend to move to usage, agent, or outcome models by 2030, with seat-based share falling from 21% to 15%. The teams winning are pricing on what the product does, not how many logins it has.
The era where every B2B company could count on price increases as a reliable growth lever is at least partially ending, unevenly, one exposed tier at a time.
