What’s the Endgame for SaaS Pricing Models After the AI panic? (Spoiler: Outcome-Based Isn’t It)

After late January 2026, the SaaSpocalypse felt imminent. AI agents and vibe-coding tools would make per-seat SaaS pricing models obsolete before most vendors had alternatives ready.

The panic has since “cooled”, but the underlying question remains. A Cruxy survey of 300 SaaS CEOs in April 2026 found 97% plan to retire seat-based pricing within two years. Yet 94% said seat-based pricing currently aligns with their product’s value.

In my opinion, this contradiction reveals the reality of pricing in 2026: hybrid pricing over “outcome-based”.

Outcome-based pricing has moved from a fringe idea to a mainstream narrative, but most implementations remain closer to output-based than outcome-based. Meanwhile, hybrid pricing blends a predictable per-seat fee with credit-based charges for AI features and has become the most common model.

This article covers the shape SaaS pricing is taking in 2026, the main SaaS pricing models that will “survive” (spoiler: most of them 🤷‍♀️), and a practical guide to choosing the right model for your product.

The state of SaaS pricing after the AI panic

The “SaaSpocalypse” sell-off in early 2026 was extreme, but market analysts noted it was partly correct. Booz Allen CTO Bill Vass told Tech Brew: “Agentic systems will disintermediate these SaaS vendors and these ERP vendors. The market just realized that, and that’s why you see the big reaction.” AI agents can now read business policies and generate software custom-fit to your needs, undermining the SaaS premise that standardized platforms beat custom builds on cost and reliability.

The companies most at risk are those with thin moats. Faisal Masud, who heads HP’s digital and lifecycle services, told Tech Brew that smaller solutions for niche use cases are most at risk compared to larger platforms (especially after renewal negotiations).

However, when looking at those larger platforms, you see that Salesforce reported 12% year-on-year revenue growth and $41.5 billion in full-year revenue for FY2026. ServiceNow saw subscription revenue climb 21% in Q4 2025. These platforms are still growing, even though you can theoretically “clone” their products with Claude’s Fable 5 (at double the cost of Opus 4.8, by the way).

Tamar Yehoshua, chief product and AI officer at Atlassian, said: “SaaS is not going away. People are going to have to build software. People are going to have to communicate with other people.”

In the end, the “SaaSpocalypse” was less of an extinction event and more of a structural reshape of how SaaS should be priced.

Seat-based-only pricing is the most at-risk

The fundamental problem with seat-based pricing today is one of unit economics. When one AI agent can do the work of ten human users, the per-user model loses its relationship to value. As Satya Nadella framed it, seats are becoming “just entitlement to some consumption.”

Deloitte’s 2026 TMT Predictions explicitly flagged this: AI agents don’t appear in an admin’s license dashboard, and their work doesn’t map to the pricing unit at all.

Carrie Osman, founder and CEO of Cruxy, described the pattern: “Pricing decisions are typically based on gut feel, broad competitor benchmarks, or segment averages, with no granular, quantified view of headroom at the customer, product, or market level.” The result is that loyal customers are undercharged, high-potential segments are overlooked, and seat-based renewals are contested by procurement teams seeking AI efficiency gains. Osman noted, “Static seat-based models were always going to become extinct. Agentic AI is just accelerating the timeline.”

Coming back to Salesforce as an example, its operations ran 380,000-plus customer support interactions using its own Agentforce agents, with 84% fully resolved without human intervention.

Sales engineers reported a 10% reduction in customer service seats across 90 enterprise accounts due to AI making agents more efficient. In short, Salesforce’s own AI product model is cannibalizing its own per-seat revenue, trading seat-based expansion for “outcome-based” growth.

Where pricing is actually landing: The outcome-based hybrid era

Surprisingly, the market is not leaning toward outcome-based pricing, as Salesforce and Intercom did. Instead, hybrid pricing combines a predictable base subscription with variable consumption on top. According to Kyle Poyar’s 2026 State of B2B Monetization survey of over 230 software companies, 37% use hybrid as their primary structure, the most common option by far.

The PricingSaaS 500 Index tracked more than 1,800 pricing and packaging changes among the top 500 B2B and AI companies in 2025 alone, an average of 3.6 changes per company for the year.

This report found that credit models became the defining pricing innovation of 2025, with companies using them growing 126% year-on-year (from 35 to 79). This shift happens because credits sit between seat-based access and pure outcome pricing, offering more transparency than legacy licenses and more feasibility than charging purely for results (because defining and charging for outcomes isn’t possible for most products).

HubSpot, Figma, Adobe, Cursor, and Lovable all added credit models without killing the seats. The PricingSaaS Q2’s trends report also added Notion, ClickUp, and Dialpad to this list.

Additionally, hybrid is where most of the investors’ money is going: when Poyar asked which pricing model investors favor, only 5% said seat-based and 10% said flat-fee subscription. Investors favored hybrid (35%), outcome-based (26%), and usage-based (24%).

So as it turns out, seat-based pricing isn’t dead; it’s morphing into a hybrid model. It became the fastest-growing model because it combines the consistency of seats with the expansion opportunity AI usage creates.

The main SaaS pricing models (and what makes each one tick)

Most discussions of SaaS pricing models present each option as a set-in-stone choice. The reality is messier: 29% of companies now offer multiple pricing models simultaneously, up from 21% a year ago (per Poyar’s survey). The rise of hybrid pricing doesn’t make individual models irrelevant but makes understanding them more important.

Here are the pricing models we currently have:

Tiered subscription pricing

Subscription pricing, sometimes called flat-rate pricing, charges a fixed recurring fee for access to the product, typically monthly or annually. Most SaaS companies add tiers that offer different feature sets at multiple price points.

The psychological appeal of tiered pricing is the decoy effect. When offering three plans, the middle option tends to convert best because the top and lower tiers are priced to make the middle look more cost-effective. Buffer, Zoom, and most project management tools still use tiered pricing as their primary model because it makes revenue predictable and provides a natural upsell path as customers’ needs grow.

Compared to usage-based, tiered pricing’s drawback is that it’s harder to drive revenue expansion. A team deriving ten times more value from a tool pays the same as one deriving one-tenth, so upgrading a plan needs a better reason than “more usage.”

Seat-based pricing

Per-user pricing charges by the number of individuals with access to the software (there’s also a less common variation called “per-active-user pricing”, which charges only for users who log in during the period). In this model, value (and thus revenue) scales with the number of people using the product.

As I mentioned, seat-based isn’t dead, but combining it with a credit-based model is becoming more common as AI features launch. Seat-based prices are no longer sold in isolation but complemented by usage-based billing to expand revenue. This lets clients use AI features to optimize seat count without hurting the product’s revenue growth.

Feature-based pricing

Feature-based pricing (sometimes called per-feature pricing) charges customers based on which capabilities or modules they activate rather than how many people use them.

Amazon Web Services is a classic example. It charges separately for compute, storage, and data transfer, each priced at the granular service level. Enterprise software vendors have used variants of this model for years by selling base platforms with paid add-ons for premium functionality.

This model helps buyers control what they pay for, preventing them from paying for features they don’t use. However, there’s a risk of cognitive overload if the menu of options is too long or the pricing logic is too opaque.

Usage-based pricing

Usage-based pricing (also called pay-as-you-go or consumption-based pricing) charges customers based on how much they consume, such as API calls, data processed, messages sent, transactions completed, and so on. You see it with Stripe’s per-transaction fee, Twilio’s cost per text message/phone call, etc.

The appeal of this pricing model is that it aligns directly with value delivered, making unit economics transparent for both buyer and seller. Customers pay only for what they use, which also lowers the barrier to adoption and helps you close clients who would hesitate to commit to annual contracts.

That’s why we’re seeing a wave of usage-based SaaS (combined with seat-based). Maxio found that 83% of AI-native SaaS companies offer usage-based pricing, largely because the underlying cost structure (token consumption) maps naturally to a consumption-based model. A Metronome’s research showed that 77% of the largest software companies use consumption pricing specifically to unlock revenue expansion from existing customers.

The challenge for this model is revenue forecasting. Usage varies, and finance teams at vendors and customers struggle to build reliable budgets around highly variable monthly bills. This is why usage-based pricing works well with subscriptions.

Freemium

The freemium pricing model offers a permanently free product with limited features or usage, while having unlimited premium plans for teams with advanced use cases. Products like Canva, Notion, Slack, and Dropbox, for example, all built their user bases substantially through freemium.

The model is a growth mechanism that lowers acquisition costs and creates product usage habits before asking users to pay. It works because users invest time setting up the product, building workflows, and importing data, making leaving feel “harder” due to Cialdini’s commitment principle.

However, this model rarely works in B2B SaaS. Freemium conversion rates typically range from 2% to 5%, so free products usually need a large user base to grow. Common advice is to use a reverse trial (full-feature access for a limited period, then defaulting to the free tier) to capture the conversion rates of a regular free trial and the user volume of the freemium model.

Outcome-based pricing

Outcome-based pricing charges customers for results rather than access. You pay when the product delivers a defined outcome, whether that’s a resolved support ticket, a booked sales meeting, or a processed loan application. The model has attracted serious attention because companies like Salesforce are using it to incentivize the adoption of their AI features.

Intercom’s Fin is a known example. Starting at $0.99 per resolved conversation, Fin grew from $1M to $100M-plus ARR on this model and backed the pricing with a $1M performance guarantee if resolution targets weren’t met. Sierra AI, as another example, takes the model further by charging only when its agent resolves an issue without any human intervention, and hit $150M-plus ARR in early 2026 from that position alone.

The clear challenge for most companies is that while “resolution” in customer support is measurable, most AI use cases lack a discrete metric for tracking “outcomes.” To consider this model, Kyle Poyar coined a framework called CAMP:

• Consistency of outcomes across diverse customer environments.
• Attribution of results clearly to the product.
• Measurability of the outcome in real time.
• Predictability of what customers will spend.

Choosing the right pricing model for your SaaS product

There is no universally correct pricing model. As Kyle Poyar puts it: “The best you can do is choose the pricing model that allows you to tell the unique story of what you do, who you’re for, and why you’re better than the alternative — and have a plan to manage the inevitable downsides.”

What follows are the pricing fundamentals I think matter the most during this SaaS reshaping:

Aim for value-based pricing

Value-based pricing is the approach that sets a price based on the customer’s perceived value and willingness to pay. When executed well, it’s the highest-margin approach available and the best way to withstand commoditization pressure.

To apply it, you need a “value metric”, which is the metric you use to track the success a customer gets from your product (e.g., transactions processed, active users, resolved conversations). If prices scale with this metric, your product will naturally expand revenue as customers succeed. If pricing is disconnected from value, some customers pay less than the value they get, causing revenue churn.

Value-based pricing is also the hardest to implement correctly. Most prices are set on gut feel or benchmarking, without understanding what customers find valuable. But for value-based, you must prove the value your product delivers before you can charge for it. That means customer research, segmented usage data, and a credible framework connecting product activity to business outcomes.

For early-stage companies, I recommend starting with cost-plus or competitor-based pricing. Once you collect data, pay close attention to how your product delivers value to different customers, then make pricing changes as you solidify your hypotheses. Jumping aggressively to value-based pricing can cause high churn and damage early customer relationships.

Mind your company’s stage

Early-stage companies with less than $5M ARR are still running a product discovery process alongside a sales process. They’re learning which customers get the most value, which use cases create stickiness, and which features predict long-term retention. A simple flat subscription or tiered plan keeps sales cycles short, removes pricing friction, and generates clean usage data for smarter pricing decisions.

Kyle Poyar’s report found 37% of early-stage companies use flat fees, not because they’re optimal, but because they close faster and create fewer objections. Adding usage-based prices before having enough data to calibrate usage limits or predict consumption is one of the most expensive early-stage pricing mistakes.

As I suggested above, you first need to develop strong evidence of consistent value delivery, reliable tools to measure consumption, and sufficient financial runway to absorb revenue unpredictability. Only then can you consider adding other pricing models with your value metric in mind.

Make enterprise pricing more predictable

If your company offers enterprise packages, you know enterprise procurement operates on annual budget cycles. Finance teams need a number they can commit to in Q4 for an expense that begins in Q1.

As a result, adding variable pricing (whether usage-based or outcome-based) creates planning problems for buyers who need to lock down software spend twelve months in advance.

Some companies address this by pairing a minimum commitment (a prepaid block of outcomes, credits, or consumption) with overage pricing for volume above that floor. This strategy preserves the incentive of usage-based pricing while giving procurement teams a number they can approve. It also protects vendor revenue from the volatility in pure consumption models.

Another challenge for enterprise products is adding pricing that supports expansion without renegotiating the entire contract. Flat pricing with an expansion cap can cause customer lifecycle stagnation in B2B SaaS. If your pricing model doesn’t naturally expand as customers get more value, every upsell becomes a negotiation instead of a usage milestone.

Consider segmenting your pricing to protect expansion revenue

A single pricing model rarely optimizes all customer segments simultaneously. Different segments have fundamentally different value structures, budget processes, and expansion ceilings.

Segmented pricing creates conditions for healthy net revenue retention (NRR). It lets each segment expand naturally within a pricing structure tailored to their usage patterns.

The best starting point is to segment your customer base by usage, industry, or company size, then measure average expansion revenue per account (ARPU) by segment. If your highest-value segment is also your lowest-growth one, the pricing model isn’t capturing its value, and you need to iterate on it.

Treating pricing as a product discipline

The companies with the healthiest pricing in 2026 are not the ones that got pricing right on day one. They’re the ones that built the operational muscle to change pricing quickly, learn from the results, and iterate without breaking customer trust. The PricingSaaS 500 data makes this concrete: the top companies averaged 3.6 pricing changes in a single year, and Lovable made roughly one meaningful pricing update per month while scaling to $200M ARR.

Jason Lemkin’s framing at SaaStr captures the mindset shift: “Ship multiple options, let customers self-select, measure what works, and iterate fast. The companies that win the next five years won’t just have the best AI. They’ll have the pricing architecture that captures its value as it grows.”

This discipline requires version control on pricing decisions, clear ownership, and the willingness to run experiments that might not work.

My recommendation is to A/B test different packaging options, price points, or trial lengths across distinct customer cohorts. It will generate empirical pricing data that internal debates never will, and guide your next pricing decisions following the value metric.

Validating your pricing assumptions with customer and product data

You must make pricing decisions based on product usage data. Understanding which features drive upgrade decisions, where users face friction, or which usage patterns predict long-term retention are all questions that require instrumentation.

In-app surveys are one of my favorite methods. For instance, you can trigger a survey for users who’ve just hit a paywall or downgraded to ask about what would make them upgrade.
The answers to this question are not always the price point itself. Sometimes it’s the absence of a specific feature in the lower tier, or the timing of the limit’s wall in their workflow. Those qualitative insights are what help you understand the reason behind a price change and make you more likely to improve your free-to-play conversion rates.

For teams moving toward usage-based or outcome-based pricing specifically, proper data analysis is non-negotiable. You need to define what your pricing metric means in your product context, build the tracking to measure it reliably, and validate that customers agree with your definition of “value” before you attach a price to it.

Userpilot’s Product Analytics and Agent Analytics, for instance, give product teams the data layer to support this, from tracking which features drive upgrade intent to measuring AI agent conversation outcomes and self-service resolution rates.

Pricing is a competitive question now, not a configuration decision

The SaaSpocalypse accelerated a structural shift that was already in motion. SaaS vendors are now pressured to build a pricing architecture that captures the value their product creates as it grows, rather than one that made sense in 2019 and got left on autopilot.

So if you want to understand how your customers actually experience your product (where they stall, what triggers upgrades, what your usage data says about willingness to pay), Userpilot gives your product team the tools to find out. Book a demo to see how in-app surveys, analytics, and agent tracking connect product usage to the pricing decisions that matter most.