How to Price a SaaS Product: A 7-Step Pricing Framework (2026)

Most founders treat pricing as a single number to test. It is actually a system, and the number is the least important part of it.

Why SaaS Pricing Is a Growth Lever, Not a Guess

Pricing is one of the few levers that touches every stage of the revenue funnel at once. Get it wrong and you feel it everywhere: in the top of the funnel, in expansion revenue, and in the churn report.

The three outcomes price controls: acquisition, expansion, retention

A pricing model shapes who even considers buying (acquisition), how much revenue grows from existing accounts without new sales effort (expansion), and whether customers stick around once the initial excitement fades (retention). A metric that is too aggressive scares off self-serve buyers. A metric that is too generous caps expansion and leaves money on the table as customers grow. Firms like OpenView Partners and Bessemer Venture Partners have built entire benchmarking practices around the idea that pricing architecture, not the price itself, is what separates compounding SaaS businesses from stalled ones.

Why cost-plus fails for software

Cost-plus pricing (take your cost to serve, add a margin) makes sense for physical goods with a marginal cost per unit. Software has a marginal cost near zero after the first customer, so cost-plus either underprices value delivered to power users or overprices a basic plan relative to what it costs you to run. Value-based pricing, anchored to what the product is worth to the buyer, is the standard approach recommended across pricing literature, including work published by Harvard Business Review on pricing strategy.

The real cost of underpricing early

Underpricing early feels safe because it removes friction from the first sale. The cost shows up later: a price increase on an existing base is far harder than raising prices for new customers only, and an early price point becomes an anchor your own sales team has to argue against years later. It is easier to discount down from a defensible price than to raise up from an accidental one.

Step 1: Choose Your Value Metric (The Decision That Matters Most)

The value metric is the unit you charge for: seats, API calls, contacts, GB processed, revenue managed, or an outcome like deals closed. Everything else in this framework sits on top of this choice.

What a value metric is vs. a feature

A feature is something the product does. A value metric is the unit that scales with the value the customer receives. Confusing the two is the most common pricing mistake: locking core functionality behind a tier (a feature decision) is not the same as choosing what you charge per unit of usage (a metric decision). You can gate features for packaging purposes, but your value metric should track value delivered, not an arbitrary feature boundary.

Per-seat vs. usage vs. outcome-based metrics

Per-seat pricing is predictable and easy to understand but can penalize adoption (teams avoid adding users to control cost). Usage-based pricing, the model popularized by infrastructure and API-first companies, scales naturally with consumption but makes budgeting harder for the buyer. Outcome-based pricing (charging on results like revenue recovered or leads generated) aligns incentives tightly but requires an attribution story the customer trusts.

How to test whether your metric scales with customer value

Plot your top accounts' usage of the candidate metric against their actual expansion and satisfaction. If your largest, happiest customers are not also your highest-usage accounts on that metric, it is not tracking value, and you should test an alternative before building pricing infrastructure around it.

Step 2: Pick a Pricing Model That Fits the Metric

Once you know what you are charging for, decide how you charge for it. The model is the structure; the metric is the unit inside it.

Flat-rate, tiered, per-user, usage-based, and hybrid models

When usage-based pricing wins (and when it scares buyers)

Usage-based pricing wins when consumption correlates directly with the value delivered and when buyers can forecast their own usage reasonably well. It scares buyers when bills are unpredictable month to month, which is why many usage-based vendors add committed-use discounts or spending caps. Research from Zuora on subscription business trends has tracked the steady rise of usage and hybrid models as SaaS has matured beyond simple seat licensing.

Matching the model to your buyer's budgeting process

Enterprise buyers with annual procurement cycles want predictable, pre-approved line items, which favors tiered or per-seat models with annual contracts. Developer and technical buyers who expense smaller amounts monthly tolerate usage-based variability far better. Match the model to how the budget actually gets approved, not just to what feels fair.

Step 3: Research What Buyers Will Actually Pay

Guessing a price point is the fastest way to leave money on the table or price yourself out of deals you could have won.

The Van Westendorp price-sensitivity method

The Van Westendorp Price Sensitivity Meter asks respondents four questions: at what price is the product so cheap you'd question quality, a bargain, getting expensive, and too expensive to consider. Plotting the responses reveals an acceptable price range rather than a single guessed number. Survey platforms including Qualtrics document the method as a standard tool for pricing research.

Willingness-to-pay interviews vs. surveys

Surveys scale but surface what people say they would pay. One-on-one willingness-to-pay interviews, where you walk a prospect through real scenarios and watch where they hesitate, surface what they would actually do, and tend to catch objections a survey never would.

Mining existing customer and churn signals

If you already have customers, your own data is a research instrument. Which accounts expanded without being asked? Which churned citing price? Which negotiated hardest? Pricing researchers at firms like ChartMogul, which builds subscription analytics for SaaS companies, point to expansion and churn-reason data as leading indicators that are often ignored in favor of outside benchmarks.

Step 4: Package Into Tiers Buyers Can Self-Select

Tiers turn a single price into a menu, letting different buyers self-select into the plan that matches their needs and budget.

Good-better-best and the decoy tier

Three tiers (good, better, best) is the most common structure because it gives buyers a comparison point without overwhelming choice. A well-placed middle tier, priced and featured so most buyers land there, is often called the decoy: it exists partly to make the top tier look reasonable by comparison.

Which features to gate vs. keep in the base

Keep in the base plan whatever is required for a new user to get real value fast; gate what scales with team size, usage volume, or advanced workflows that only mature customers need. Gating core value too early kills activation; gating nothing limits upgrade reasons.

Designing the upgrade path to drive expansion

Every tier boundary should map to a moment a growing customer actually hits: adding a tenth teammate, exceeding a usage cap, needing an integration or compliance feature. If customers never bump into a limit, you have no natural expansion trigger and are relying entirely on new logo sales for growth.

Step 5: Set Your Actual Price Points

With a value metric, a model, willingness-to-pay research, and a tier structure in place, you finally have enough information to attach real numbers to each plan.

Anchor each tier against the one next to it

Buyers judge price by comparison, not in isolation. The ratio between tiers matters as much as any single number: a jump from the entry tier to the middle tier that roughly doubles the price while meaningfully expanding what is included reads as fair, while a jump that triples the price for a marginal feature increase reads as a trap. Set the price gaps between tiers deliberately, using the willingness-to-pay range from Step 3 as the outer bounds.

Set the entry price and the ceiling first, then fill the middle

Start with the two numbers you have the most signal on: the lowest price that still filters for a serious buyer, and the highest price your best current customers (or the top of your Van Westendorp range) will bear. Once those two ends are fixed, the middle tier's price is a design choice, not a guess, since it only has to sit credibly between them and capture the bulk of your target segment.

Build your discounting policy before you need one

Decide your annual versus monthly discount, your maximum discretionary discount for sales, and who is allowed to approve anything beyond it, before the first negotiation forces an ad hoc answer. A discount handed out once becomes the number a customer expects at every renewal, so treat your discount ladder with the same rigor as the sticker price itself.

Step 6: Launch, Grandfather, and Communicate Changes

How you roll out a pricing change often matters more to retention than the change itself.

Rolling out to new vs. existing customers

The lowest-risk path is applying new pricing to new customers only, leaving existing contracts untouched until renewal. This avoids a retroactive price shock and gives you a live comparison between old and new pricing performance.

When and how to grandfather legacy plans

Grandfathering (letting existing customers keep their current price or plan) preserves trust but can leave you running multiple pricing structures indefinitely. A common middle ground is grandfathering for a fixed window (for example, through the next one or two renewal cycles) with advance notice before the new pricing applies.

Writing the price-change announcement

A good price-change notice explains what is changing, when, why in terms of value added (not just cost), and exactly what the customer needs to do, if anything. Silence or a surprise invoice line is what triggers support tickets and cancellations, not the price change itself.

Step 7: Measure and Iterate (Pricing Is Never Done)

Pricing is not a launch task you finish once. It is a system you monitor the same way you monitor product usage.

The metrics to watch: ARPU, net revenue retention, win rate

Subscription benchmarking work from groups like SaaS Capital and the annual surveys referenced by McKinsey's growth insights consistently treat net revenue retention as one of the clearest signals that pricing and packaging are working, or are not.

How often to revisit pricing

Most mature SaaS companies review pricing annually at minimum, and revisit sooner after a major product expansion, a new competitor entering with a different model, or a shift in the target buyer (for example, moving upmarket from SMB to enterprise).

Running a safe price test

Test new pricing on new customers or a specific segment (a new geography, a new acquisition channel) before rolling it out broadly. This isolates the effect of the price change from existing-customer dynamics and gives you a clean read on win rate and deal size impact before you touch your whole base.

Frequently Asked Questions

How much should I charge for my SaaS product when I'm just starting out? Start from willingness-to-pay research and comparable products in your category rather than a round number that feels safe. Price low enough to get initial traction but high enough that a price increase later is a modest adjustment, not a shock.

What is a value metric and how do I choose one? It is the unit you charge for (seats, usage, outcomes). Choose the one that scales most closely with the value your best, most successful customers actually get from the product, which you can check by comparing usage of the candidate metric against expansion and satisfaction among your top accounts.

Is usage-based pricing better than per-seat pricing? Neither is universally better. Usage-based pricing aligns cost with value for consumption-heavy products but creates bill unpredictability; per-seat pricing is predictable but can discourage adoption. Match the model to how your buyers budget and how closely usage tracks value.

How many pricing tiers should a SaaS product have? Three is the common default (good, better, best) because it gives buyers a comparison without overwhelming them. Add a fourth, typically an enterprise tier with custom pricing, once you have buyers who need negotiated contracts, compliance features, or dedicated support.

Should I offer a free trial or a freemium plan? Free trials suit products that sell themselves quickly once tried and are backed by active sales or onboarding. Freemium suits products with low marginal cost per free user and a clear, naturally occurring trigger (a usage cap, a team-size limit) that converts free users to paid over time.

How often should I change my SaaS pricing? Review pricing at least annually, and revisit sooner after a major product change, a shift in your target buyer, or evidence from the metrics in Step 7 (declining ARPU, falling net revenue retention, rising price-cited churn) that current pricing is not working.

How do I raise prices without losing existing customers? Apply new pricing to new customers first, grandfather existing customers for a defined window, give advance notice before any change takes effect, and frame the announcement around added value rather than cost increases alone.

Conclusion

The number on your pricing page is the last decision in this framework, not the first. Get the value metric and packaging right in steps 1 through 4, and the actual price points in step 5 become a narrower, lower-stakes choice. Treat step 7 as permanent: the companies that compound revenue fastest are the ones that keep measuring and adjusting pricing long after launch, not the ones that got it perfect on day one.