How to Do Market Validation Before Building a B2B SaaS Product
Positive feedback is not market demand. A problem that exists is not necessarily one customers will pay to solve. Here is the five-step process to validate before you build.
Don't build your way into a market
A founder has an idea for a SaaS product. The problem sounds real. The solution seems technically feasible. A few people in their network say it would be useful. The natural next step is to start building. I think this is where many early-stage founders make an expensive mistake.
I've spent years working on B2B technology marketing, product positioning and outbound GTM. One lesson keeps coming back: you can have the right product idea and still fail because you misunderstand the buyer, the problem or the market.
When I developed an email warm-up solution under my previous agency, I assumed customers wanted better domain reputation and higher open rates. But conversations with prospects revealed a different priority — they didn't want to manage another tool. They wanted someone accountable for email deliverability outcomes.
Define the problem and your ideal customer
Before researching keywords, conducting interviews or launching an outbound campaign, write down one problem and one ICP. Start with one problem, one ICP and one hypothesis.
Strong: "Revenue operations teams at B2B SaaS companies struggle to ensure contractual price increases are applied to every eligible subscription on time."
Define the ICP before defining the market size
An ICP describes the type of company most likely to experience the problem and have a reason to solve it. For a B2B SaaS product, define:
IndustryWhich sector is most likely to experience the problem?
Company sizeDoes the problem become more significant at a particular scale?
Technology environmentDoes the company use a system or workflow relevant to the problem?
BuyerWho experiences the problem, owns the process or controls the budget?
TriggerWhat event makes the problem urgent enough to act on?
Current alternativeHow does the company handle the problem today?
The same product, a different ICP — real results
The product hadn't changed. The market focus had. A different ICP can produce a materially different response to the same underlying product.
One clearly defined ICP, one problem statement and a list of assumptions you need to test.
Talk to 10–15 target buyers
The purpose is to uncover patterns, understand current behaviour and identify what you don't yet know. Don't start by pitching your idea. Start by understanding their world.
Ask about what they already do
Avoid hypothetical questions — they invite polite encouragement
A record of buyer interviews, the language customers use, existing workarounds and a shortlist of problems worth investigating further.
Not sure how to structure your market validation outreach?
We help early-stage B2B teams build problem-led outbound campaigns that generate real signal — not just polite responses.Check demand signals beyond Google search
Customer interviews tell you what a small group of people experience. Demand research helps you understand whether there are broader signals worth pursuing. Use two complementary starting points: search behaviour and conversations happening in the market.
Combine signals rather than relying on one
| Search volume | Conversations | What to investigate |
|---|---|---|
| High | High | Is there a specific underserved segment or unmet need? |
| Low | High | Are buyers discussing a problem without searching for a recognised category? |
| High | Low | Are the searches relevant to your ICP, or mostly general research? |
A demand-research summary with relevant keywords, trends, buyer discussions, competitor alternatives and unanswered questions.
Test how much the problem hurts
A problem can be real without being important enough to justify a new product. Assess its frequency, consequences, urgency and the effort buyers already invest in solving it. Look for evidence of behaviour, not stated interest.
Measure the pain through five dimensions
A ranked list of problems supported by evidence of frequency, consequences, current workarounds and priority.
Validate with targeted outbound
Interviews and desk research can leave a critical question unanswered: will people outside your immediate network respond to a specific description of the problem? This is where targeted outbound becomes a market-validation tool, not just a lead-generation channel.
Lead with the problem, not the product
Strong: "We're researching how SaaS teams manage subscription price increases across Salesforce and billing systems. When subscriptions have different indexation dates, how do you ensure eligible price changes are applied on time?"
A case study: when the first ICP was wrong
In one email-verification infrastructure engagement, the founder initially targeted mature email-verification companies. We tested it with ~40 accounts — the hypothesis didn't hold. Most saw the offering as a competitor, not a vendor.
We pivoted to two new segments — early-stage verification companies (~50 accounts, 5 confirmed the problem) and mature data-technology companies (~100 accounts, 7 showed genuine interest). The second finding was particularly useful because the founder had initially assumed these companies would not be interested. The market had changed since his earlier experience.
This is why I don't treat validation as a campaign with a predetermined lead target. It is a sequence of hypotheses that can be confirmed, rejected or refined.
A record of what you tested, who responded, what buyers independently confirmed and which assumptions you should change.
Go, adjust or stop — how to read your results
Market validation should lead to a decision, not an endless request for more research. Before you start, define what evidence would make you proceed, revise the hypothesis or stop investing in the idea.
Multiple independent confirmations
Buyers independently describe the problem, consequences matter, and there is credible evidence of willingness to invest in a solution.
Signal from a different angle
The problem exists, but strongest responses come from a different ICP, narrower use case or different description. Refine and retest.
No meaningful problem found
Repeated tests fail to uncover a meaningful problem, existing workarounds are adequate, or the problem consistently ranks below more urgent priorities.
Ready to test your market hypothesis with real outbound?
We build and run market-validation outbound campaigns for early-stage B2B SaaS founders — so your first signal comes from real buyers, not your network.Five market-validation mistakes to avoid
Asking friends and family
People who know you may want to encourage you. Their feedback can be useful, but it shouldn't substitute for conversations with independent target buyers.
Treating survey answers as buying intent
Someone saying they would use your product is not the same as someone changing their behaviour, allocating a budget or committing to a next step.
Relying only on AI-generated research
AI can help generate hypotheses and organise findings. It cannot establish that a specific group of buyers experiences a problem simply by producing a convincing explanation of it.
Validating the idea but not the price
A problem can be real and still not support a viable business. Investigate what customers currently spend, who controls the budget and what alternatives compete for the same money.
Stopping after one positive signal
One enthusiastic buyer may have an unusual need or a personal relationship with the founder. Look for independent confirmation across multiple companies.
How to use AI in market validation without fooling yourself
AI can make validation faster, but it should support evidence gathering rather than replace it.
Market-validation checklist
Before building your B2B SaaS product, make sure you can answer these questions:
Problem and ICP
Customer discovery
Demand research
Outbound validation
Commercial viability
Key takeaways
- Positive feedback from your network is not the same as market demand — test with independent target buyers.
- Define one ICP and one problem hypothesis before doing anything else. Mixing two ICPs produces noisy, unactionable signal.
- Ask about past behaviour, not hypothetical enthusiasm — "would you use this" invites politeness, not buying intent.
- Combine search signals and LinkedIn conversations — neither alone tells the full story.
- A strong validation signal is a buyer who independently describes the problem, names a consequence, and can point to what they currently spend to address it.
- Validation does not end when someone agrees the problem exists — test the solution, the buying process and the price before building.
