How to Price a SaaS Product in 2026: Models, Steps and Examples

Short answer: How to price a SaaS product: pick a pricing model that matches how customers get value from it (per seat, per usage, flat tiers, or a hybrid), anchor the number on the value customers receive rather than on what the product costs you to run, package it into two to four tiers with a clear reason to upgrade, then test the price with real buyers and adjust.

Your costs set the floor below which you lose money. Your competitors set the context customers will judge you against. But the price itself should come from the value — the time, money or revenue your product creates for the customer. Everything in this guide is a way of working that out, step by step.

Pricing is the SaaS decision founders most often make by gut feel and then avoid revisiting. That is understandable: pick a number that is too high and you worry nobody will buy; pick one that is too low and the business quietly cannot fund itself. The way out of that trap is to treat your first price as a hypothesis, not a verdict. A hypothesis can be tested, and this guide ends with how to test it.

If you are still deciding what to build, work through the idea first — our list of 150+ profitable SaaS ideas is a good starting point — because the right price depends heavily on who the product is for and what it replaces.

The main SaaS pricing models

A pricing model is the structure of what you charge for. Choosing one is really choosing your value metric — the thing that scales as the customer gets more value: users, usage, projects, messages, storage, or simply one flat fee. The model matters because it decides who finds your price fair, how revenue grows as customers grow, and how complicated your billing becomes.

Flat-rate pricing

One product, one price, everyone pays the same. Flat-rate is the simplest model to explain, buy and administer, and it works well when customers are similar to each other and use the product in roughly the same way. It breaks down when customers differ a lot: heavy users get a bargain while light users subsidise them and quietly churn. Many products start flat and outgrow it.

Per-user / per-seat pricing

You charge for each user on the account, usually per month. Per-seat pricing fits products where each additional person genuinely gets their own value — collaboration tools, CRMs, communication software — because the customer’s team size and the value they receive grow together. It fits badly when only one person does the real work and everyone else just views results: charging per viewer feels like a tax and discourages the sharing that makes the product spread inside a company. A common softening is to charge for editors or creators and let viewers in free.

Tiered pricing

Two to four plans at different price points, separated by features, limits, or both. Tiered pricing is the most widespread structure in SaaS because it lets different customer types self-select: small customers start cheap, growing customers upgrade as they hit limits, and large customers pay for the controls and support they need. Tiers work best when the step between plans is a natural growth moment — running out of projects, seats or automations — rather than an arbitrary paywall. Packaging tiers well is a skill of its own, covered in the packaging section below.

Usage-based pricing

Customers pay for what they consume: API calls, emails sent, records processed, minutes transcribed. Usage-based pricing aligns your revenue with the customer’s activity, lowers the barrier to starting (a light user pays very little), and grows automatically with successful customers. Its weakness is predictability — customers dislike bills they cannot forecast, and your own revenue becomes harder to project. The billing unit you choose becomes part of your product’s identity: Zapier charges per task, Make per credit and n8n per execution, and our Zapier vs Make vs n8n comparison shows how three tools selling similar automation end up with very different bills for the same workload purely because of the unit each one meters.

Freemium

A free plan sits alongside paid tiers. Freemium is best understood as a marketing channel rather than a pricing model: the free tier exists to acquire users at scale and convert a portion of them later. It works when the product spreads naturally (users invite users), when free users cost you very little to serve, and when there is a clear, felt moment where paying becomes obviously worth it.

It fails when free users are expensive to support, when the free tier is so generous that nobody needs to upgrade, or when there is no built-in reason for users to bring in other users. For many early products, a time-limited free trial tests willingness to pay faster and more honestly than a permanent free plan, because everyone on a trial is there to evaluate buying.

Hybrid pricing

A base subscription plus a usage or per-seat component on top — for example, a platform fee that includes a starter allowance, with overage charged per unit beyond it. Hybrids combine the predictability of a subscription with the growth of usage-based pricing, which is why so many mature products drift towards them. The trade-off is complexity: the customer has to understand two charges, and you have to bill both accurately. For a first product, start with the simplest model that fits; you can add a hybrid layer later when you know which usage actually predicts value.

Pricing models compared

Model	You charge for	Fits best when	Watch out for
Flat-rate	One fixed fee for everyone	Customers are similar and usage is uniform	Heavy users underpay, light users overpay
Per-seat	Each user on the account	Every user gets independent value (collaboration, CRM)	Charging for viewers discourages sharing
Tiered	Feature/limit bundles at set prices	Customers differ in size and needs	Tiers separated by arbitrary paywalls instead of growth moments
Usage-based	Units consumed (calls, messages, records)	Usage tracks value closely and varies widely	Unpredictable bills make buyers nervous
Freemium	Nothing for the free tier; tiers above it	The product spreads user-to-user and free users are cheap to serve	Free users who never convert but still cost support
Hybrid	Base fee + usage or seats on top	You need a predictable floor plus growth with usage	Billing complexity and customer confusion

Three ways to set the number — and which one should lead

Once you have a model, you still need the actual figure. There are three classic approaches to finding it, and they are not equal. The strongest pricing uses all three, but in a specific order: value sets the price, costs set the floor, and competitors set the context.

Cost-based pricing (your floor)

Add up what it costs to serve a customer — hosting, third-party APIs, payment processing, support time — and mark it up. Cost-based pricing is easy and feels safe, and it is genuinely useful as a floor: below it, every customer you win loses you money. But as a method for choosing the price it is weak, because your costs have almost nothing to do with what the product is worth to the buyer. Software also has low marginal costs, so cost-plus arithmetic tends to produce prices far below what customers would happily pay for a real problem solved. Use costs to check a price, not to create one.

Competitor-based pricing (your context)

Look at what alternatives charge and position yourself relative to them. This is worth doing — customers will do it, so you should know the landscape, including the free alternative of doing the task manually or with a spreadsheet. But copying a competitor’s number copies their cost structure, their customer segment and their strategy, none of which are yours. A rival may be pricing low to win a market you are not chasing, or high because their product does more. Treat competitor prices as the frame your price will be judged inside, not as the answer.

Value-based pricing (your anchor)

Price according to the economic value the product creates for the customer: time saved, costs removed, errors avoided, revenue gained. This is the approach worth leading with, because it aligns what you charge with what the buyer actually cares about, and because it usually supports a higher, more defensible price than cost-plus.

The practical method is estimation, not precision: work out what the problem currently costs the customer (their time at a reasonable rate, the tools or staff they pay for today, the cost of mistakes), estimate the share of that value your product realistically delivers, and price at a fraction of it — low enough that the customer’s return is obvious, high enough to fund a real business. The exact fraction is a judgement call you refine through the research in the next section. The principle is what matters: a customer who saves ten hours a month is not buying your hosting bill, and your price should not pretend otherwise.

How to price a SaaS product, step by step

Step 1: Research what customers will pay

Before setting a number, gather evidence about willingness to pay. The cheapest reliable method is structured conversations with the people you intend to sell to — the same customers you would interview to validate the idea itself. In those conversations, a classic research tool is the Van Westendorp Price Sensitivity Meter, which asks four questions about a product:

1. At what price would it be so cheap that you would doubt its quality?
2. At what price would it feel like good value?
3. At what price would it start to feel expensive, but still worth considering?
4. At what price would it be too expensive to consider at all?

Plotting the answers across a group of respondents shows a range of acceptable prices rather than a single magic number — which is exactly what you need at this stage. Treat the output as a range to test, not a guarantee: people are imperfect at predicting what they will pay, so combine it with the other signals in this guide. Just as useful are simpler probes: what do you pay today to deal with this problem, what did you last pay for a similar tool, and what would need to be true for this to be an easy yes?

Step 2: Choose your value metric

Decide the one thing that scales with value — seats, projects, messages, records, locations — because that choice quietly determines your model, your tiers and your upgrade path. A good value metric is easy for the customer to understand, grows as they get more value, and is hard to resent. If customers cannot predict what drives their bill, they will distrust it; if the metric does not grow with their success, your revenue will not either. Many products keep one primary metric and use features, not metrics, to separate the higher tiers.

Step 3: Map features and limits to tiers

List everything your product does, then sort it into what everyone gets, what growing customers need, and what only larger or more serious customers value. Core functionality belongs in every paid tier — a product that does not work properly until the third plan makes the first plan a trap. Higher tiers earn their price through scale (higher limits, more seats), depth (advanced features, integrations, automation), and assurance (priority support, security controls, uptime commitments, onboarding). The packaging section below goes deeper on how to draw these lines.

Step 4: Set the actual numbers

Start from the value research, not from your costs. A practical sequence: set the price of the tier you expect most customers to choose first, anchored on the value that segment receives; then price the entry tier low enough to be an easy start without undercutting the main tier’s appeal, and the top tier high enough to reflect the extra scale and assurance it carries. Keep the gaps between tiers meaningful — small gaps give no reason to upgrade, enormous gaps strand customers in between. And do not be afraid of the number feeling slightly uncomfortable: founders systematically anchor on what they would pay, and you are not your customer.

Step 5: Sanity-check against costs and margins

Now use the floor. Estimate your cost to serve one customer on each tier — infrastructure, per-unit API or processing costs, payment fees, and a realistic slice of support — and check that each tier clears it with room to spare. Software economics only work if revenue per customer comfortably exceeds the cost of serving them, because the surplus is what funds development, support and acquiring the next customer. If the entry tier only works when customers barely use the product, its limits are drawn in the wrong place.

Also check the upgrade maths: when a customer grows into the next tier, does your revenue grow faster than the extra cost they impose? If not, the metric or the limits need rework.

Step 6: Test the price before you commit to it

A price you have not tested is a guess wearing a suit. Practical tests, cheapest first: put the price in sales and validation conversations and watch reactions (hesitation about price is information; instant acceptance can mean you aimed low); publish a pricing page and measure how many visitors start a trial or join a waitlist at that price; run the same offer at two price points to different slices of an audience where that is feasible and fair.

Pricing tests slot naturally into the wider validation process — our guide to how to validate SaaS ideas covers interviews, landing pages and pre-sales in detail. Whatever you learn, remember that the first price is version one. Plan to revisit it as the product and your customer mix change.

Packaging: how many tiers, and what goes in each

Three tiers is the most common packaging pattern in SaaS, popularised as Good-Better-Best: a genuinely usable entry plan, a fuller plan aimed at your typical customer, and a top plan for customers who need scale or assurance. Two tiers can work when your customers are very uniform; four can work when you clearly serve distinct segments. Beyond four, most buyers struggle to compare, and comparison fatigue costs sales.

Whatever the count, give each tier a job:

• Entry tier — removes the risk of starting. Full core product, modest limits, standard support. It should be good enough that customers succeed and grow into the next tier, not so stripped that they fail and blame the product.
• Middle tier — built for your typical customer, and the tier your pricing page should make the natural default. This is where the features that regular, growing users need live: higher limits, key integrations, automation, team features.
• Top tier — priced for scale and assurance rather than features alone: the highest limits, advanced security and admin controls, priority support, onboarding help. Customers buying this tier are often buying risk reduction as much as functionality.

Differentiate tiers on things customers can feel — limits they can hit, capabilities they can name, response times they can rely on — rather than on long feature checklists nobody reads. And place your gates at growth moments: a customer should meet the upgrade prompt because their usage grew, which feels like success, not because an ordinary feature was hidden behind a wall, which feels like a trick.

Worked example: pricing a hypothetical product

Everything in this section is a hypothetical illustration. The product, the people and every number are invented to show the method — they are not data about any real business.

Imagine PulseBoard, a fictional tool that turns a small marketing agency’s scattered ad and analytics accounts into one client-ready monthly report. Its founder works through the steps:

Value first. In interviews, agency owners say a single client report takes a junior staff member roughly five hours a month to assemble by hand. PulseBoard’s founder estimates the tool cuts that to about half an hour of review. Even valuing that staff time cautiously, each client report the agency produces carries a real labour cost that the tool mostly removes — and the agency can either pocket the saving or serve more clients with the same team. That saving, per agency, is the value pool the price should be drawn from.

Costs as the floor. The founder estimates serving one agency account — data connections, processing, storage, payment fees and a share of support — costs a few dollars a month at expected usage, and an order of magnitude less than the labour value above. The floor is low; the pricing decision will be made by value, exactly as it should be.

Metric and tiers. The natural value metric is client workspaces: the more clients an agency reports on, the more time the tool saves. The founder packages three tiers — Starter at $19/month for up to 3 client workspaces, Growth at $49/month for up to 15 workspaces with scheduled automatic sending, and Studio at $119/month for unlimited workspaces, white-label reports and priority support. Growth is the tier aimed at the typical agency; its price sits well below the monthly labour cost of producing even a handful of those reports manually, so the return is easy for a buyer to see, while the per-workspace economics still improve for PulseBoard as agencies grow.

Testing. Before launch, the founder shows the pricing page to agency owners from the validation interviews. Two say Growth is cheaper than expected — a signal worth noting, not an instruction to raise prices on the spot. The founder launches at these numbers, watches which tier people choose and where they stall, and schedules a pricing review after the first cohort of paying customers has lived with the product. That review, not launch-day anxiety, is where the next pricing decision gets made.

Common SaaS pricing mistakes

• Pricing from costs alone. Cost-plus arithmetic produces a floor and calls it a price. You end up cheap, busy and unable to fund growth, selling to customers who chose you because you were cheap.
• Copying a competitor’s number. Their price encodes their costs, segment and strategy. Undercutting it by a little gives you their strategy with worse margins.
• Underpricing to be safe. Low prices do not just leave revenue behind; they set expectations, attract the most price-sensitive customers, and are painful to raise later. Nobody ever complained their way into a sustainable business by being the cheapest.
• Too many tiers. Five plans with overlapping features make buyers do homework instead of buying. If you cannot explain the difference between two tiers in one sentence, merge them.
• Gating the wrong features. Locking basic functionality behind higher tiers makes the entry plan feel broken. Gate scale, depth and assurance — not the core job the product exists to do.
• Per-seat pricing when value is not per-seat. If one person configures and twenty people glance at a dashboard, per-seat billing punishes adoption. Charge the role that gets the value.
• Discounting instead of diagnosing. “Too expensive” usually means “I don’t see the value yet” or “I’m not the right customer”. A discount fixes neither; it just trains buyers to negotiate.
• Set-and-forget. The price you launch with reflects a product and a market that will both change. A price reviewed on a schedule improves; a price left alone decays.

When and how to raise prices

Prices should rise when the value does. Reasonable triggers: the product has gained substantial capability since the price was set; your customers have shifted towards larger or more demanding segments; your costs to serve have genuinely increased; or your testing shows buyers accepting the current price without the hesitation that suggests you are near the ceiling. A price that was right for a rough first version is almost certainly wrong for the mature product it becomes.

How you raise matters as much as how much:

• New customers first, or everyone at once — decide deliberately. Applying new prices to new signups first lets you confirm the market accepts them before touching existing relationships. Moving everyone at once is simpler and treats customers equally, but concentrates the reaction into one moment.
• Give existing customers notice and a reason. Explain what has improved and what the change funds. Silence plus a higher charge is the combination that reads as disrespect.
• Consider grandfathering. Letting existing customers keep their current price — permanently, or for a defined period — rewards loyalty and removes most of the churn risk from the change. It costs less than it looks, because those customers were acquired at the old price’s economics anyway.
• Pair the change with visible value. A price rise that lands alongside a feature or improvement customers have asked for is a different conversation from a bare increase.
• Revisit the packaging, not just the numbers. Sometimes the right move is not “+20% everywhere” but new limits, a new top tier, or a different metric. Structure problems are not solved by arithmetic.

Frequently asked questions

How much should I charge for my SaaS product?

There is no universal number, because the right price is a fraction of the value your specific customers receive — and that varies enormously. Work it out in this order: estimate what the problem costs your customer today, decide what share of that value your product delivers, price below that value so the customer’s return is obvious, then check the result clears your cost to serve with room to spare. If you need a starting point for the conversation, the worked example above shows the full method applied to a hypothetical product.

What is the best pricing model for a new SaaS product?

The simplest one that matches how your customers get value. For most new products that means flat-rate or a small set of tiers, because they are easy to explain, easy to bill and easy to change. Add usage-based or hybrid elements later, when you have real usage data showing which activity actually predicts value and retention. Complexity in pricing should be earned by evidence, not adopted because bigger companies have it.

Should I offer a free plan or a free trial?

They test different things. A free trial tests willingness to pay: everyone on it is evaluating a purchase, and the deadline forces a decision. A free plan tests distribution: it only pays off if free users bring in other users, cost very little to serve, and meet a clear moment where paying becomes obviously worthwhile. If your product does not spread user-to-user, a trial usually tells you more, faster. Whatever you choose, decide what the free offer is for before you build it.

How do I know if my price is too low or too high?

Read the reactions, not just the conversion rate. Consistent signs of underpricing: buyers accept instantly without negotiating, prospects say the price is lower than expected, and customers on your highest tier are using far more than it was designed for. Signs of overpricing: qualified buyers who understand the product still walk away citing price, sales cycles stall at the pricing page, and discounting becomes routine. Both are hypotheses to check with the willingness-to-pay research in Step 1 — a handful of structured conversations beats months of guessing.

What is value-based pricing, in plain terms?

Charging according to what the outcome is worth to the customer, instead of what the product costs you to deliver. If your tool reliably saves a business several hours of paid work every month, a price that represents a modest share of that saving is easy for the buyer to justify and sustainable for you. Costs still matter — they set the floor you must clear — but under value-based pricing they stop being the ceiling.

How often should I change my SaaS pricing?

Review it on a schedule — many teams look at pricing whenever the product ships a major capability change, and at least periodically even when nothing forces the question — but change it only when the review produces evidence: new value delivered, a different customer mix, cost changes, or test results showing the current price is off. Frequent small changes confuse customers and complicate grandfathering; never reviewing at all guarantees the price drifts out of line with the product.

Conclusion

Learning how to price a SaaS product is mostly learning what to ignore. Ignore your hosting bill as a source of prices — it is a floor, not a strategy. Ignore the urge to copy or undercut a competitor whose economics you cannot see.

Instead, build the price from the customer’s side: choose the model and metric that track the value they receive, package tiers around their growth, set the number as a defensible share of that value, check it clears your costs, and test it with real buyers before treating it as final. Then keep it under review, because the product, the market and the customer will all move — and a price that moves deliberately with them is one of the few advantages a small SaaS can compound year after year.