Zurück

SaaS Market Research: Methods, Questions and Metrics

Matthieu Saussaye

SaaS market research is the work of finding out which buyers have a problem painful enough to pay for every month, what they will pay, and why they stay or leave. Unlike one-off product research, it never ends: a subscription business has to re-earn the purchase at every renewal, so research runs before launch, during growth and at every churn event.

The short version:

  • Start with the ideal customer profile (ICP). Company size, role, tech stack and trigger event matter more than demographics. Most SaaS research fails because it surveys "users" instead of the people who sign the contract.

  • Validate the problem before the product. Problem interviews with 15 to 30 target buyers tell you whether the pain is real, frequent and currently budgeted.

  • Test willingness to pay with structured methods such as Van Westendorp, Gabor-Granger or MaxDiff on features, not with "would you pay $X?".

  • Measure product-market fit with the Sean Ellis question once you have active users, and track the share who would be "very disappointed" without you.

  • Run win-loss and churn interviews continuously. They are the cheapest competitive intelligence a SaaS company can buy.

  • Pair surveys with product analytics. Behavior shows what users do; research explains why.

SaaS research methods by stage

The right method depends on where the company is. A pre-seed founder and a Series C product team are answering different questions, with different samples available to them.

Stage

Key question

Best methods

Sample you can reach

Output

Pre-product

Is the problem real and budgeted?

Problem interviews, secondary research, search demand analysis

15 to 30 target buyers recruited from LinkedIn, communities or a B2B panel

Problem statement, ICP draft, kill or continue decision

MVP / beta

Does the solution fit the workflow?

Usability tests, concept tests, beta feedback surveys

Beta users and waitlist sign-ups

Prioritized feature list, onboarding fixes

Early revenue

Do we have product-market fit, and at what price?

Sean Ellis PMF survey, Van Westendorp, Gabor-Granger, win-loss interviews

Active users (ideally 40+ responses per segment), recent wins and losses

PMF score by segment, price range, positioning

Growth

Which segments and features drive expansion?

MaxDiff, conjoint, segmentation survey, NPS with follow-up

Customers, trial users, broader market via panel

Packaging tiers, segment priorities

Scale

Why do customers churn and competitors win?

Churn interviews, brand awareness tracking, market sizing

Churned accounts, category buyers

Retention playbook, share-of-mind trends

Notice that the sample changes at every row. Early on you have almost no one to ask, so depth beats volume. Later you have thousands of users, and the risk flips: you only hear from people who already chose you.

Step 1: Define the ICP and the buying committee

B2B SaaS rarely has one buyer. A typical purchase involves a user who feels the pain, a manager who owns the budget, IT or security who can veto, and sometimes procurement. Research that only talks to end users will overestimate demand, because users are not the ones weighing your price against everything else on the budget.

Before writing a single question, write down for each role:

  • Firmographics: industry, employee count, region, revenue band.

  • Technographics: the tools they already pay for and must integrate with.

  • Trigger events: a new hire, a failed audit, a funding round, a tool sunset. Triggers predict buying better than company size.

  • Current workaround: spreadsheets, an incumbent tool, an agency, or nothing. The workaround is your real competitor.

Then decide which roles each study needs. A pricing study should include budget holders. A usability study should not.

Step 2: Validate the problem with interviews

Problem interviews come first because they are cheap and they kill bad ideas early. Talk to 15 to 30 people who match the ICP, and talk about their past behavior, not your idea. People are poor predictors of what they would buy, but reliable reporters of what they did last quarter.

Good problem-interview questions:

  1. "Walk me through the last time you had to [do the job]. What happened?"

  2. "What did you use to get it done? What did that cost you in time or money?"

  3. "What is the most frustrating part of that process?"

  4. "Have you looked for a better way? What did you try, and why did you stop?"

  5. "Who else was involved in deciding how this gets done?"

  6. "If this problem disappeared tomorrow, what would change for you?"

Stop and rethink if fewer than a third of interviewees have actively tried to solve the problem. A pain nobody has spent money or effort on is usually not a pain anyone will subscribe to fix.

Step 3: Size the market from the bottom up

Top-down figures from analyst reports describe the whole software category and say little about your slice. For SaaS, bottom-up sizing is more useful and easier to defend to investors:

  • Count the accounts that match your ICP (from company databases, registries or LinkedIn filters).

  • Multiply by the share that has the problem, estimated from your interviews or a screener survey.

  • Multiply by a realistic annual contract value from your pricing research.

This gives a serviceable addressable market you can rebuild every time the ICP changes. Keep the assumptions in a visible table so anyone can challenge them.

Step 4: Test willingness to pay

Asking "would you pay $49 a month?" produces polite yeses. Use a method that forces trade-offs:

Method

What you ask

What you learn

Watch out for

Van Westendorp

Four price questions: too cheap, a bargain, getting expensive, too expensive

An acceptable price range and an indifference point

Works best when respondents understand the product; show a clear description first

Gabor-Granger

Would you buy at this price? (price varied across respondents or rounds)

A demand curve and the revenue-maximizing price point

Stated intent overshoots real conversion, so read it as relative, not absolute

MaxDiff

Of these features, which matters most and which least?

A ranked importance score per feature

Keeps features in or out of tiers; does not set price on its own

Conjoint

Choose between packages that vary on price, features and limits

The value of each feature in currency terms

Needs larger samples and careful design

For B2B, ask pricing questions to people who hold budget, and anchor on the unit you actually charge for: seats, usage, workspaces or outcomes. A per-seat price and a per-usage price can test very differently with the same audience.

Step 5: Measure product-market fit

Once you have active users, the most widely used survey benchmark comes from Sean Ellis. As First Round Review describes, it asks users "how would you feel if you could no longer use the product?", and after benchmarking nearly a hundred startups Ellis found that companies with strong traction almost always exceeded 40% answering "very disappointed" (First Round Review).

Run it on users who have experienced the core value, typically people who used the product at least twice in the last two weeks. Then segment the results. A 25% overall score can hide a 50% score in one segment that should become your new ICP. Follow up with:

  • "What type of people do you think would most benefit from this product?"

  • "What is the main benefit you receive?"

  • "How can we improve the product for you?"

Our full guide to product-market fit covers how to act on the score.

Step 6: Run win-loss and churn research

Every closed deal and every cancellation is a free research participant. Win-loss interviews within two to four weeks of the decision show how buyers compare you with alternatives. Churn interviews show where the value broke down.

Useful questions for lost deals and churned accounts:

  1. "What problem were you trying to solve when you started evaluating tools?"

  2. "Which other options did you seriously consider?"

  3. "What was the deciding factor in the end?"

  4. "At what point did you start to doubt we were the right fit?"

  5. "What would have had to be true for you to stay?"

  6. "What are you using now, and what is it doing better?"

Code the answers into themes and count them each quarter. A rising share of "missing integration" or "too expensive for our size" is an early signal you will not see in a dashboard until renewals drop.

SaaS research KPIs worth tracking

Research should connect to the numbers the business already watches. These are the metrics SaaS research most often explains:

Metric

What it tells you

Research that explains it

Activation rate

Share of sign-ups who reach the first value moment

Onboarding surveys, usability tests

Logo and revenue churn

Share of customers or revenue lost per period

Churn interviews, exit surveys

Net revenue retention

Expansion minus contraction and churn in existing accounts

Feature importance (MaxDiff), packaging research

PMF score

Share of engaged users who would be very disappointed without you

Sean Ellis survey with follow-up probing

Net Promoter Score

Willingness to recommend

NPS survey with open follow-up on the reason

Win rate

Share of qualified deals won

Win-loss interviews, competitive perception

See how to run NPS properly, including why the open follow-up question is worth more than the score.

Sampling and recruitment constraints in SaaS

B2B SaaS populations are small and busy. That shapes every study:

  • Small universes. If your ICP is "heads of finance at 200 to 2,000 employee companies in one country", the whole population may be a few thousand people. You will not get 1,000 completes, and you do not need them for directional decisions. Use the sample size calculator to see what precision a realistic sample buys.

  • Verification. B2B panels contain people who overstate their job title to qualify for higher incentives. Screen with knowledge questions (which tools they use, what their team does) rather than one self-reported title question.

  • Incentives. Professionals expect more per minute than consumers. Short, relevant surveys with a clear reason to answer beat generic gift cards.

  • Your own users are biased. Customers already chose you. For market-level questions such as awareness or category demand, add non-customers from a panel. Our guide on finding survey respondents lists the options.

  • Privacy. Interviewing users of a B2B product often means handling business contact data under GDPR or similar rules. Get consent, state the purpose, and avoid pasting customer confidential data into research tools that train on it.

Common SaaS research mistakes

  • Surveying only happy users. In-app surveys reach active users. Churned accounts and lost deals hold the harder truths.

  • Treating feature requests as demand. A request from one large account is a sales signal, not a market signal. Weight requests by segment and by how often the underlying problem appears.

  • Asking hypothetical intent. "Would you use this?" answers are inflated. Ask about past behavior and current spend.

  • Pricing research with the wrong respondent. End users anchor on their personal tools; budget holders compare against headcount and alternatives.

  • One-off studies. SaaS markets move every quarter. A light, recurring program beats a big annual report that is stale by the time it ships.

Run SaaS interviews at survey scale

The bottleneck in SaaS research is usually depth: problem, churn and win-loss interviews are the most useful studies and the hardest to run on more than a handful of people.

SmartInterview runs those conversations as AI-moderated surveys by voice or text. The AI follows up when an answer is vague ("too expensive compared to what?"), works in the respondent's language, and codes every open answer into themes, so 200 churn interviews come back as a ranked list of reasons instead of a transcript pile.

Start free

Frequently Asked Questions

What is SaaS market research?

SaaS market research is the process of studying target buyers, competitors and pricing for a subscription software product. It covers problem validation, ICP definition, willingness to pay, product-market fit measurement and continuous churn and win-loss research, because a SaaS company must win each customer again at every renewal.

How do you do market research for a SaaS startup?

Start with 15 to 30 problem interviews with people who match your ideal customer profile, then size the market bottom-up from account counts. Once you have a prototype, run concept or usability tests, and once you have active users, measure product-market fit with the Sean Ellis survey and test pricing with Van Westendorp or Gabor-Granger.

How many interviews do you need to validate a SaaS idea?

Usually 15 to 30 well-recruited interviews per target segment are enough to see whether a problem is real and recurring. If the same pains stop appearing new after about 10 interviews in a segment, you have reached saturation; if every interview tells a different story, your segment is too broad.

What is a good product-market fit score for SaaS?

The common benchmark is 40% of engaged users saying they would be "very disappointed" if they could no longer use the product. It comes from Sean Ellis's survey of nearly a hundred startups, as reported by First Round Review. Read it by segment, not only overall.

How do you research SaaS pricing?

Use structured willingness-to-pay methods rather than asking if a price is acceptable. Van Westendorp gives an acceptable range, Gabor-Granger gives a demand curve, MaxDiff ranks features for packaging, and conjoint values each feature. Ask budget holders, and test the pricing unit you actually charge for.

What is the difference between SaaS and technology market research?

Technology market research covers hardware, infrastructure and one-time software purchases. SaaS research focuses on recurring revenue: activation, retention, expansion and churn, so it runs continuously and leans heavily on existing-customer and lost-customer research.

Table of Contents

Verwandte Artikel

Kostenlos anmelden

Kostenlos anmelden

Kostenlos anmelden