Because AI agents can do jobs without human help, selling software by the user seat is dead. If vendors keep charging per employee, helping customers get more efficient will only shrink their own sales. To survive, they must start billing for actual usage or guaranteed business results. This shifts the financial risk onto the software developer, who now has to run automatic audits to prove the AI actually delivered results before getting paid.
Software companies must stop selling seats and start selling finished work
Updated
Backlinks
- Workday Q1 FY2027: Agentic AI Doubles Adoption, Flex Credits & Sana Expand Operating Margins
Workday is successfully maintaining operating margins by introducing Flex Credits, transitioning its monetization away from traditional per-seat licensing to consumption-based pricing for autonomous agents.
- The Shift from Seat-Based to Outcome-Based AI SaaS Pricing: Vendor Playbooks and Procurement Realities
It documents HubSpot's programmatic transition of its Breeze AI agents from monthly charges to pay-per-result and resolved-conversation pricing models.
- The AI Agent GTM Shift: The Backlash Against Outcome-Based Pricing and the Rise of Consumption and Hybrid Models
It analyzes the practical friction and backlashes occurring as software transitions away from traditional seat licensing to outcome and consumption models.
- The Great SaaS Reset: Outcome-Based and Hybrid AI Agent Pricing in 2026
This finding showcases Zendesk's implementation of outcome-based pricing alongside a double-verification process to solve the verification challenge of proving AI successfully resolved an issue.
- AI Agent Consolidation Thesis: Most Startups Will Merge or Be Acquired by 2028
A lasting business advantage doesn't come from the cleverness of the AI itself, which anyone can eventually copy, but from weaving your tool so deeply into a company's complex, highly specific daily work routines that they can never easily rip you out.
- Sources
When a platform gets too expensive or bloated with features, it makes creating a custom, specialized alternative look easy by comparison, driving users to ditch the standard for their own local solutions.
- SaaS Pricing Models Under Structural Siege: Enterprise Customers Pay More for the Same Tools
This indicates that traditional software vendors are actively transitioning their pricing models away from seat-based licenses in favor of billing based on consumption and usage.
- Double-Digit SaaS Inflation and Budget Volatility Drive Aggressive Procurement Tactics
SaaS companies now grow far more by maximizing the revenue they get from each user through clear proof of value and optimized pricing than they do by simply boosting top-of-funnel conversion rates.
- The Pragmatism of Ownership: The Real Economics of the $48K Home-Brew GPU Server
Paying as you go through renting or data tolls acts as a constant mental tax that makes you play it safe, while owning the hardware or native platforms creates an urge to get your money's worth that drives you to explore aggressively.
- Capital Allocation and FCF Divergence: Free Cash Flow Collapse vs. shareholder Returns
In a capital-intensive industry shift, you generate superior returns by selling the mandatory tools that everyone must pay to use, rather than being the entity that funds the massive physical infrastructure needed to host them.
- Casca (Cascading AI) Named to 2026 Best Places to Work in Fintech as AI-Native Lending Gains Traction
When established companies lay off technical staff to afford renting ready-made AI services, they accidentally feed those skilled workers directly to specialized startups that use them to rebuild old industry workflows from scratch.
- McDonald's Q1 2026: Value Menu Overhaul and McCafé Expansion Drive Modest Traffic Growth
Moving from open-ended discounts or unlimited access to a clearly defined, metered model forces customers to assign an explicit value to their consumption, which creates the constraints necessary to cultivate higher commitment rather than driving users away.
- Southeast Asia Fintech Venture Capital: Singapore Hub, AI and Embedded Finance Themes
Since value is now created by fitting directly into a user's existing workflow rather than offering a standalone tool, specialized fintech and SaaS products must become built-in features of dominant platforms to survive the collapse of the best-of-breed model.
- Google's Antigravity 2.0 and the Gemini CLI 'Bait and Switch'
When autonomous tech threatens to bypass an established platform, the platform defends itself by turning its open connections into paid tollgates, forcing the very agents that would make them obsolete to pay for access.
- The Drew Houston Era Ends: Dropbox's AI Pivot and the 'SaaS Apocalypse'
Standalone software only survives because building and integrating tools is hard, so once AI or platform integrations make doing it yourself cheaper than the subscription cost, customers will replace those standalone tools with their own builds or built-in features.
- Kailera Therapeutics IPO: $625M Raise, First New GLP-1 Public Challenger
Both show that how quickly money moves, whether rushing into pre-revenue speculation or fleeing from fears of obsolescence, is driven by high-level market stories rather than the actual performance or fundamental viability of the underlying assets.
- PLG Benchmarks 2026: The Flywheel Metrics That Separate Elite SaaS from the Rest
When the entry point that brings in users—like PLG free trials or basic GLP-1 efficacy—becomes a common commodity, survival shifts from raw, volume-focused acquisition to deep activation, requiring you to embed into workflows or offer specialized, multi-attribute benefits to overcome growth plateaus.
- Agentic AI Market Size and Growth: $800M ARR and 29,000 Deals for Salesforce's Agentforce in Q4 FY26
This shows Salesforce shifting its metric of success and billing away from user seats toward 'agentic work units' that measure the actual work completed by autonomous AI.
- Anthropic Deepens Wall Street Ties: CAIS Integrates Claude via Model Context Protocol (MCP) Alongside Pre-Built Agents and Platform Partnerships
Both findings show that forcing a mandatory data format on isolated, custom fields acts as a bridge, opening up hard-to-move, locked-away assets so they can be ingested by fast, standard systems.
- Zendesk Bets on Outcome-Based Pricing — Direct Challenge to Seat-Based SaaS
Zendesk is abandoning traditional software pricing that charges per human license in favor of billing customers only when its AI agents successfully resolve support issues. This ties their revenue directly to the completion of specific jobs rather than system seat counts.
- Zendesk's Outcome-Based Pricing: Shifting to 'Verified Resolutions' and Multi-Tier Billing to Solve Customer Friction
Both fields use a separate transaction step to bypass legacy friction, holding off on syncing the actual value and payment until the activity meets a specific standard of quality or certainty.
- AI COGS Problem: SaaS Gross Margins Compress from 80% to 52-65% Range
To keep costs from spiraling out of control, you can't treat every task the same, meaning you have to save your expensive, high-quality resources for the few really complex problems and use cheap, good-enough approximations for the vast majority of everyday work.
- AI Application Layer Companies Hit $100M ARR in 7 Quarters — Compressing GTM Velocity
Technologies scale at extreme speeds not by making the old way slightly more efficient, but by using a new underlying process to completely wipe out the root cost itself—like obesity or human labor—which lets them skip the usual hurdles to adoption.
- Pace Raises $46M Series B Led by Thrive and Sequoia to Scale Agentic Insurtech Operations
An AI agent's usefulness and competitive edge do not scale with its intelligence alone, but are strictly limited by how deeply it is embedded in the software where work actually happens, meaning the winners will be the companies that control the workflow rather than those with the smartest models.
- The Seismic Shift to Outcome-Based AI Pricing: CCaaS and CRM Restructure to Mitigate Seat Compression Fears
This line explicitly contrasts the elimination of traditional seat-based licensing with a transition to charging only for specific, completed tasks like resolved issues and qualified leads.
- AI-Native Revenue Operating Systems Are Replacing Fragmented GTM Stacks
Moving raw employee know-how into permanent software allows companies to treat workers as interchangeable while keeping their expertise locked in, which explains why laying people off is actually a deliberate move toward building intelligence directly into their technology.
- Enterprise Platform AI Agent Strategies: Zendesk, ServiceNow, and Workday
Zendesk is abandoning the traditional per-seat SaaS model to bill customers exclusively when its AI software successfully resolves a customer support ticket.
- ServiceNow: AI Disruption Fears and Turnaround Momentum Driven by Strategic Pivot to Non-Seat "Assists" and Volumetric Pricing
When classic moats like patents or per-seat pricing lose their value, established companies must stop charging for basic inputs and instead charge for the actual AI-driven work that gets completed in a business process to keep making money.
- Incumbent Agentic Pricing: How Salesforce, HubSpot, and Atlassian Defend Per-Seat Revenue Models
Both systems are moving from requiring live human approval to relying on pre-authorized, automated code, which means they must use built-in safeguards like payment retries or outcome-based credit pools to keep revenue steady without human oversight.
- Enterprise Case Studies: Autonomous Agents Delivering Measurable ROI in 2026
Autonomous agents scale not through better reasoning, but by operating in environments where the underlying software has been broken down into tiny, easily discoverable APIs.
- How AI-Native ERP Startups Are Killing the Per-Seat and Implementation Fee Models
Both drug companies and SaaS providers are being forced to abandon high-friction pricing models, like high drug list prices and per-seat software licenses, because regulation and AI automation are transforming those old revenue anchors from legitimate value-capture into indefensible rent-seeking.
- Cognizant and Accenture: AI Services Narrative Splits Between "Fox in the Hen House" and "Validation"
An optional innovation becomes a mandatory requirement when you rebrand it from a simple efficiency tool into a prerequisite for basic operational or physiological stability.
- AI-Native ERPs Deploy Automated Migration Engines to Challenge Legacy Giants, but Face Onboarding Friction
New challengers beat entrenched giants not because their final product is better, but because they eliminate the exact switching friction—like data migration or talent recruitment—that previously locked customers in.
- Platform Wars Heat Up: Agentic AI Moves from Copilots to Autonomous Resolution
To displace a horizontal incumbent, you stop competing on breadth and instead force your specific technical constraints to become the mandatory environment where the work is performed, effectively making your specialized architecture the prerequisite for achieving the desired outcome.
- Figma: Re-accelerating to 46% Growth and Proving the Viability of AI Credit Monetization in Q1 2026
Switching from unlimited, instant delivery to scheduled, constrained delivery turns unpredictable customer usage into high-value, budgeted commitments, proving that clear boundaries are actually what make a service sticky for enterprise clients rather than a barrier to growth.
- 8x8 Accelerates Usage-Based Pivot as AI Handles More Customer Interactions
This line shows that 8x8 is transitioning its pricing model away from billing per human seat because automated AI interactions are doing the work instead.
- Zendesk Launches Outcome-Based Pricing and Autonomous AI Agents at Relate 2026
Zendesk is billing customers only when its AI successfully settles a customer support ticket instead of charging per employee seat. This directly shifts their revenue model from human logins to the actual volume of work completed by its digital agents.
- The Death of Flat-Seat SaaS: The Shift to Outcome-Based and Pooled Consumption Pricing
When innovations become highly effective, providers must shift from flat or volume-based fees to outcome-linked pricing, capping the cost of the service based on the actual medical savings or task efficiency it generates so the purchase remains budget-neutral for the buyer.
- Salesforce Q1 FY2027: Agentforce ARR Passes $1B, Headless 360 & MCP Open-Source Adoption Reshape CRM
Instead of selling logins for human seats, Salesforce is pricing its platform based on the credits and web calls consumed by autonomous software agents. This directly links their billing model to the volume of automated tasks these digital workers perform.
- Salesforce's Agentic Enterprise License Agreement (AELA): The "All-You-Can-Eat" Trap
In both cases, adopting a transformative but unproven intervention is temporarily cushioned by cheap upfront pricing that hides the actual unit costs, eventually forcing the payer to either find massive operational efficiencies or face a sudden collapse in affordability.
- Brazil's Pix Automático: Redefining Recurring Payments and Cardless Subscriptions in 2026
When we switch from paying for indirect access, like seat licenses or credit card networks, to paying directly for actual results, like AI resolution or automatic account transfers, we need to build in an intentional delay to audit performance and replace the implicit trust of older systems.
- The Frontier Labs Services Land Grab: OpenAI and Anthropic Deploy Forward-Engineering arms to Capture Wall Street Workflows
Both findings show how a dominant platform vendor first encourages outside hobbyists, intermediaries, and independent integrators to build on their technology to drive adoption, only to eventually squeeze them out—using gatekeeping in the FPGA market and vertical integration in AI—to grab whatever profit margin those partners have left.
- Snowflake vs. Databricks: The Battle for the Agentic Data Layer in 2026
This tool represents the precise type of automatic auditing system software vendors must implement to prove to customers that their digital agents successfully finished a task.
- Nvidia Enters Premium PC Market with 1-Petaflop "RTX Spark" Superchip
To displace a horizontal incumbent, you stop competing on breadth and instead force your specific technical constraints to become the mandatory environment where the work is performed, effectively making your specialized architecture the prerequisite for achieving the desired outcome.
- Supreme Court Upholds Medicare Price Negotiations as Novo Nordisk Slashes Blockbuster Prices by 50%
Both industries are being forced to shift from charging for inputs to charging for results because old measurements like user seats and list prices no longer reflect the actual usefulness the buyer gets, forcing companies to tie their prices directly to proven outcomes just to stay in the market.
- UiPath Q1 FY2027: First-Ever GAAP Profitability and the Complementary Paradigm of Deterministic vs. Agentic Automation
To successfully scale enterprise AI, we have to stop letting AI run as an independent agent and instead lock generative intelligence into a rigid, governed setup, either by compiling fluid agent outputs into deterministic, auditable scripts or by housing separate agents inside a single, constraint-based operating system.
- Financial Data Platforms: Bloomberg's Agentic Network Strategy
Both findings show that successful enterprise AI adoption requires companies to stop treating the technology as a metered, autonomous, standalone utility and instead build it into their daily workflows as budget-predictable, verifiable infrastructure to overcome institutional skepticism.
- ITC Infotech and InsureMO Partner for Agentic AI Insurance Modernization in Emerging Markets
Autonomous agents scale not through better reasoning, but by operating in environments where the underlying software has been broken down into tiny, easily discoverable APIs.
- AI Agent Startup M&A Wave 2025–2026
The wave of strategic AI acquisitions and the shift toward flat-rate enterprise licensing both show that companies are no longer treating AI as a temporary tool to pay for by the use, but as permanent capital infrastructure that they need to either own outright or lock down with long-term contracts.
- Medicare GLP-1 Cost-Neutrality Analysis: MFN Pricing Still Falls Short by ~$18B
When a business switches from charging for time and effort to charging for results, it has to set up an external auditing process to legally define what counts as value, which ultimately shifts the company's focus from actually doing the work to designing the rules that measure it.
- Intuit Lays Off 3,000 (17% of Workforce) to Sharpen AI Focus
Because AI is turning proprietary workflow software from a defensible subscription business into a cheap utility, companies must either cannibalize their own legacy labor models to pivot, like Intuit, or watch their valuations permanently collapse as the software serving as their underlying collateral loses its pricing power, which is threatening private credit.
- Eli Lilly's Multi-Indication Strategy: Expanding the GLP-1 TAM via Joint Pain and Psoriatic Arthritis Trials
An optional innovation becomes a mandatory requirement when you rebrand it from a simple efficiency tool into a prerequisite for basic operational or physiological stability.
- AMD Sparks Outrage by Stripping Linux Support from Vivado's Free Tier
If a vendor can block access to a primary asset like an FPGA or AI output using their proprietary toolchain, design portability stops being an optional optimization and becomes the buyer's only way to avoid inevitable rent-seeking.
- Druid AI Production Telemetry: How Enterprise AI Agents Actually Behave at Scale
Since raw frontier AI models cannot handle complex business tasks without a highly precise interface, companies must either limit the scope of work to rigid, controlled workflows to prevent chaos, or bypass the integration hurdles entirely by using direct, hands-on human engineering to force the connections.
- The Wikipedia Crisis: Union-Busting and Corporate Professionalization at WMF
Operational intermediaries, whether internal support staff or external consultants, always end up in one of two traps: either managers squeeze them so hard for efficiency that it ruins the actual value they provide, or they get cut out completely when primary providers take over the work directly to capture those extra profits.
- CVS Caremark Restores Zepbound Coverage and Adds Foundayo Pill, Erasing Novo's Formulary Advantage
To survive disruption when high-stakes markets commoditize, established companies stop trying to win on product features and instead become non-negotiable infrastructure—whether through pharmaceutical formulary access or foundational AI-agent plumbing—forcing the market to adopt them as a default rather than a choice.
- Novo Nordisk Partners with OpenAI to Embed AI Across Drug Discovery, Manufacturing, and Supply Chain
When competition and shrinking margins turn a proprietary molecule or a raw model into a commodity, companies stop trying to sell the output itself and start making their money on the deep operational integration and infrastructure.
- insured.io Launches AI-Powered Virtual Claims Agent for Insurance Carriers
B2B software no longer wins by getting customers to evaluate a flashy AI feature, but by proving the tech can plug directly into their backend systems and vanish into their daily workflows.
- Professional Services Firms and AI: The Accenture Playbook and Market Context
Both findings show that threatened companies survive disruption not by competing with AI, but by shifting to become the essential delivery or infrastructure layer that the very AI agents meant to replace them actually rely on to function.
- Model Labs as Strategic Acquirers: Anthropic and Mistral Deepen Their Stacks
When platform providers bundle adjacent tools to force usage and lock users in, they inevitably drive those users to set up routing middleware to maintain their bargaining power, cost-efficiency, and vendor-neutrality against proprietary price hikes.
- AI Is Now Table Stakes — Integration Depth and Time-to-Value Are the Real Differentiators
Moving from selling software features to selling friction reduction explains why AI is making basic capabilities cheap to copy, forcing both buyers and sellers to focus on how much operational drag is actually eliminated instead of seat counts.