SaaS Per-Seat Pricing Is Dead: 79 Vendors Shift to AI

By the end of 2025, 79 of 500 tracked enterprise software vendors — including Adobe, Salesforce, HubSpot, and Atlassian — had quietly abandoned the per-seat pricing model that powered the SaaS industry for two decades. That's more than double the 2024 count, and it signals a structural break that every company paying for software subscriptions will eventually feel.

SaaS Per-Seat Pricing: A 20-Year Business Model, Quietly Dismantled

The per-seat model (charging a fixed annual fee per employee who accesses the software — for example, $150/user/year for a CRM tool) was the foundation of enterprise SaaS (Software as a Service — cloud tools billed on subscription) for most of the 2000s and 2010s. It was predictable for both vendor and buyer: companies knew exactly what they'd pay, vendors knew exactly what they'd earn. Annual contracts, locked headcount, locked revenue.

AI changed the math. When an AI feature can do the work of five users — or generate revenue that one user could never create alone — charging per seat stops making sense for the vendor. The new model charges based on usage: how many AI actions you run, how many documents you process, how many leads you auto-enrich. A company deploying AI aggressively pays more. A company that barely touches AI features pays less.

Kyle Poyar, a former partner at OpenView (a venture capital firm that specialized in tracking SaaS business models), monitored 500 enterprise software vendors across multiple years. His finding: by end of 2025, 79 had formally shifted to AI usage-based pricing — up from roughly 35–40 in 2024. That is a 158% acceleration in a single year, a pace that no major software vendor can afford to ignore.

Which Giants Moved — and What the AI Pricing Switch Looks Like

The four most prominent names in the shift are household brands across corporate IT departments:

• HubSpot — moved from flat per-seat CRM (customer relationship management — software that tracks sales leads and customer interactions) licensing to AI action-based billing for features like prospect research and email generation
• Adobe — introduced consumption-based pricing for AI generative features inside Creative Cloud (its design software suite), separate from standard seat access charges
• Salesforce — launched AI usage tiers for Agentforce (their autonomous AI assistant product that handles customer service and sales tasks) sitting on top of base subscriptions
• Atlassian — introduced metered billing for AI features inside Jira and Confluence (team project management and internal documentation tools), charged per AI interaction rather than per licensed user

All four companies share a critical trait: they sell to procurement teams (the people in companies who approve software budgets and sign multi-year contracts), not to individual consumers. When these four shift their revenue model simultaneously, every enterprise IT buyer is forced to rethink how to forecast software spend — often without a clear playbook for doing so.

Why Buyers Take the Hit While Vendors Pocket the Upside

Under the old model, a 200-person company buying Salesforce could budget precisely: 200 seats × $150/seat = $30,000/year, locked in at contract signing. Under usage-based AI pricing, that same company might pay $30,000 in a quiet quarter and $80,000 when a sales push ramps up and AI-generated outreach actions multiply. The vendor's revenue grows automatically with the customer's success — or their over-reliance on AI features.

This unpredictability creates compounding budget risk for finance and operations teams. Three specific problems emerge:

1. Forecasting breaks down. Annual software budgets assume fixed costs. AI usage spikes — especially during product launches, hiring surges, or seasonal campaigns — can blow through predicted spend without any visible trigger.
2. Shadow spend increases. Teams that previously self-provisioned (added new users without formal IT approval) may now generate unexpected AI usage bills across multiple tools without realizing the cumulative exposure.
3. ROI calculations shift entirely. The old question was "is this tool worth $X per user per year?" The new question is "does each AI action generate more value than it costs?" — a harder calculation that requires tracking AI outputs, not just headcount on an invoice.

The 158% year-over-year acceleration in vendors making this transition suggests the window for buyers to adapt is shrinking fast. Vendors that have not shifted yet are watching the leaders — and few will resist the revenue upside of metered AI billing for long once Adobe and Salesforce prove the model sustainable.

For a practical breakdown of how AI automation tools are reshaping workplace software economics, the AI Automation guides cover frameworks for evaluating tool spend and comparing AI-powered alternatives.

The Culture Subplot: Benson Boone at Google Cloud Next

Away from the balance sheets, a telling cultural moment surfaced this week alongside the pricing data. Twenty-three-year-old pop star Benson Boone performed at Google Cloud Next in Las Vegas at Allegiant Stadium (a 65,000-capacity NFL venue), with Weezer as the opening act. The same week, it emerged that Mark Zuckerberg had worn Boone's signature blue jumpsuit — the one from his Grammy performance — at his wife Priscilla Chan's 40th birthday party in February 2026. Zuckerberg did not merely attend a Boone concert: he cosplayed as him.

Boone has now performed at three major enterprise tech conference events: Google Cloud Next (2026, Las Vegas), Dreamforce (Marc Benioff's Salesforce annual conference, 2025, San Francisco), and multiple YouTube events. A Google Cloud spokesperson called him "a natural fit for our community." Matt Hull, a Google VP, watched the set via FaceTime with his 11-year-old daughter and described it as "what it's all about." Joanna Duff, an Accenture marketing director in attendance, tried explaining his appeal to colleagues a generation older: "What is it the kids say, 'rizz'?"

The pattern is deliberate. As AI backlash, antitrust pressure, and executive security anxiety mount — separate reporting this week found tech executives quietly hiring specialized bodyguard firms offering "intelligent, conversational" personal protection — tech conference programming has become increasingly entertainment-heavy. The spectacle is a bid for cultural legitimacy at the exact moment business model legitimacy is under scrutiny. The new SaaS pricing model is a bet on AI growth; the Benson Boone booking is a bet on cultural growth. Both are responses to the same underlying anxiety.

Three Other Signals Worth Tracking in 2026

AI Hallucinations in a Wall Street Law Firm's Court Filings

Sullivan & Cromwell — one of Wall Street's most prestigious law firms, handling billion-dollar mergers and securities cases — publicly admitted in court filings that legal documents contained AI-generated hallucinations (factual statements invented by AI language models and presented confidently as true, with no basis in actual case law or documents). This marks a significant escalation from individual attorney incidents reported in prior years. A top-tier firm's formal acknowledgment signals that AI-assisted legal drafting is now an institutional liability with direct consequences in court — not merely a careless junior associate error that can be quietly corrected.

The $300 Million Crypto Heist — and Active Laundering

A $300 million cryptocurrency theft — identified as the largest in recorded history — left perpetrators actively laundering (converting stolen digital assets through layered transactions designed to disguise their origin and make tracing impossible) the proceeds at the time of reporting. Bloomberg confirmed the laundering operation was ongoing. The incident reinforces why institutional crypto adoption continues to face structural trust barriers, even as mainstream financial firms expand digital asset exposure.

Tim Cook Exits Apple After 14 Years

Tim Cook announced his departure from Apple after approximately 14 years as CEO (2011–2026 approx.). Media framing split four ways across major outlets: cash machine (New York Times), subscriptions czar (Wired), chief AirPodder (The Verge), and empire builder (Wall Street Journal). The divergent frames reflect how genuinely complex his tenure was — Cook simultaneously maximized Apple's financial engineering (services revenue, buybacks), transformed its hardware lineup (AirPods, Apple Watch, M-series chips), and built a subscription business generating tens of billions annually. Which frame wins will depend largely on whether his successor can sustain the AI transition Cook began.

If you manage software budgets or sign off on SaaS renewals, run a pricing audit on Adobe, Salesforce, HubSpot, and Atlassian contracts before the next renewal window. A flat-fee agreement that looks stable today may be quietly renegotiated into a consumption-based tier at renewal — without a headline announcement or proactive vendor communication. The teams that audit early, define usage caps, and negotiate usage-ceiling clauses now are the ones that avoid the budget surprises that hit procurement off guard when the first variable invoice arrives.

Related Content — Get Started | Guides | More News