SaaS Pricing Frameworks Compared

If you came here looking for one named framework, here is the short version: they agree more than they disagree. The Cobloom method, the Price Intelligently method and a Simon Kucher engagement will usually land you in the same price band from three different directions. What actually separates them is how much primary research each demands and what happens when you cannot collect it.

Below is each framework, what it assumes, the data it needs, roughly how long it takes, and where it breaks. Then a table mapping framework to stage, ACV band and research budget.

1. The Cobloom value based approach

Cobloom’s framework starts from the customer’s outcome and works backwards to a price. You define the result the buyer is purchasing, quantify it in their units (hours saved, deals closed, incidents avoided), then price as a defensible fraction of that value.

What it needs: fifteen to twenty five structured customer conversations, plus enough domain knowledge to challenge what customers tell you. No survey panel, no statistical software.

Where it breaks: it is a reasoning framework, so the output is only as good as the interviews. Teams with weak interview discipline produce a value story that customers politely agree with and never pay for. It also struggles when value is diffuse across a buying committee, since the champion’s hours saved and the CFO’s cost avoided are different numbers.

Use it at seed and Series A. It is the highest return per hour of any method on this list when you have fewer than 150 paying accounts.

2. Price Intelligently and the ProfitWell value metric method

The core claim is that the value metric matters more than the price point, and most SaaS companies choose theirs by accident. The method: survey your base on feature importance and willingness to pay, segment by persona, then find the unit that scales with value for each segment.

What it needs: a survey of your customers and ideally lost prospects, usually a few hundred responses for segment level confidence. This is where most early companies stall out, because they have 60 customers and no lost prospect list.

Where it breaks: the survey inherits all the usual stated preference problems, and the segmentation only works if your personas are genuinely distinct. If two segments give you the same answers, you do not have two segments, you have one segment and a story you liked.

Choosing the metric itself is the hard part, and it deserves its own treatment. SaaS Pricing Models Compared walks through per seat, usage, tiered and hybrid structures and what each does to your expansion curve.

3. Simon Kucher design to value

Simon Kucher’s consulting method inverts the usual order: design the product around the price rather than pricing the product you built. In practice for SaaS this means deciding the packaging tiers and price points first, then deciding which capabilities land in which tier to defend those points.

What it needs: real budget. A proper engagement involves qualitative interviews, a quantitative study, competitive analysis and internal workshops, typically eight to sixteen weeks. Expect a six figure fee at enterprise scale.

Where it breaks: it is overkill below roughly $25,000 ACV, and the recommendations often assume an organisational capacity to execute a repackaging that mid sized companies do not have. The deck is excellent. The migration of 4,000 existing accounts onto new tiers is the part nobody scoped.

Use it when a pricing mistake costs more than the study. At $40M ARR, a two point margin error is $800k a year, and a $120k study is cheap insurance.

4. The OpenView usage based playbook

OpenView’s work pushed usage based pricing into the mainstream, and their annual benchmarks have been the most cited source in the category for several years. The playbook: pick a usage dimension the customer controls, allow low friction entry, and let expansion happen through consumption rather than negotiation.

What it needs: product telemetry good enough to meter accurately and bill defensibly. That is an engineering project, not a pricing project, and it is routinely underestimated at three to six months of work.

Where it breaks: revenue predictability. Finance teams hate variance, and so do your customers’ procurement teams. Usage models also punish you in a downturn, when customers reduce consumption without churning and your NRR drops without a single logo lost. Twilio’s revenue pattern is the canonical illustration of both the upside and the exposure.

5. Van Westendorp price sensitivity

Four questions, one chart. You ask respondents at what price the product would be too expensive, expensive but worth considering, a bargain, and so cheap they would question the quality. The intersections give you a range of acceptable prices.

What it needs: 150 to 400 responses and about three weeks. Cost is usually under $5,000 if you field it to your own list, more if you buy a panel.

Where it breaks: stated preference. People answer without a budget, an approval process or a competing priority, and they overstate what they would pay. Treat the output as a relative instrument. It is good at telling you segment A tolerates 40 percent more than segment B. It is bad at telling you the number.

The way to use it well is as a screen before a live test. Understanding price elasticity in SaaS helps you read the curve honestly, and how to test SaaS pricing covers turning the range into a real experiment on real traffic.

6. Conjoint and MaxDiff packaging design

Conjoint analysis shows respondents realistic bundles at different prices and infers, from their choices, how much each attribute is worth. MaxDiff asks them to pick the most and least important items from small sets. Together they are the most rigorous tools available for deciding what goes in which tier.

What it needs: a good attribute list from prior qualitative work, 300 plus qualified respondents, and a specialist to design and analyse it. Budget $25,000 to $80,000 and six to ten weeks.

Where it breaks: garbage attributes produce confident garbage. The statistical machinery is sound and will happily tell you the precise value of a feature description your customers misread. It also assumes a stable market, so in a category being reshaped by AI capabilities the answers age fast.

Worth it above $25,000 ACV. Rarely worth it below, where the total revenue at risk from a packaging error is smaller than the study fee.

7. The jobs to be done pricing ladder

Rather than pricing features, you price against the progress a customer is trying to make, and build a ladder where each tier serves a bigger job. A solo marketer’s job is publishing. A team’s job is coordination. An enterprise’s job is governance and audit.

What it needs: strong qualitative work, specifically switch interviews with recent buyers about the moment they decided to change. Twenty interviews is a workable base. Cost is mostly time.

Where it breaks: it gives you excellent tier logic and almost no guidance on the actual numbers. Pair it with Van Westendorp or a live test. It also tempts teams to build tiers around jobs no segment in their base actually has, which produces a beautiful ladder where 92 percent of revenue sits on one rung.

The upside is that the ladder makes price anchoring honest. When each tier serves a visibly different job, the enterprise tier is not a decoy, it is a real product for a real buyer.

Which framework at which stage

The honest mapping. Pick by the data you can realistically collect in the next quarter, because a framework you cannot feed produces worse decisions than a simpler one you can.

Framework	Stage fit	ACV band	Research budget	Time
Cobloom value based	Seed to Series A	Any	Under $2k	2 to 4 weeks
Price Intelligently value metric	Series A to B	$5k to $50k	$5k to $20k	4 to 6 weeks
Simon Kucher design to value	Series C plus	$25k plus	$100k plus	8 to 16 weeks
OpenView usage based	Series A to C	Any, PLG heavy	Engineering time	3 to 6 months
Van Westendorp	Any with a list	Under $25k	Under $5k	2 to 3 weeks
Conjoint and MaxDiff	Series B plus	$25k plus	$25k to $80k	6 to 10 weeks
Jobs to be done ladder	Any	Any	Time only	3 to 5 weeks

The thing no framework fixes

A bad value metric. If you charge per seat for a product whose value scales with data volume, every framework on this list will help you optimise a number that is attached to the wrong unit, and you will spend three years fighting seat minimisation while your heaviest users pay the least.

Test your metric with one question: if a customer doubles the value they get from us, does their bill go up? If the answer is no, stop reading pricing frameworks and go fix that. Willingness to pay research covers the interview structure that surfaces the right unit, and value based vs cost plus pricing explains why cost anchored thinking keeps producing the wrong metric in the first place.

What to do next

Write down your current value metric and the last date anyone tested it. If that date is more than 18 months ago, or blank, that is your project. Pick the cheapest framework on the table that your data supports, run it, then plan the rollout carefully, because changing price on existing accounts is a separate discipline from choosing one. The price increase rollout playbook covers the notice periods, grandfathering rules and comms, and the pricing change readiness checklist is the pre flight before you announce anything. The full set of related material sits under SaaS Pricing Strategy.