Expert Interview SOP

An expert interview SOP gets you a working map of an unfamiliar industry in about two weeks, by talking to 5 to 8 people who see many customers from the outside. A customer knows her own problem. An accountant who serves 80 restaurants knows which problems all of them share, what they already tried and who signs the cheque. Experts cannot prove demand, since they will not buy your product, and they still save you months of wrong guesses about how the market works and hold the introductions you need.

What is the expert interview SOP and when should you use it?

An expert interview is a 30 to 45 minute call with someone who has a wide view of your customers' world: an adviser, supplier, distributor, association leader, regulator, journalist, community moderator or a former employee of a company in the space. You ask about structure and history: how buying works, which rules apply, what failed before and why.

Use this SOP:

• When you enter an industry you have never worked in, such as healthcare, logistics, education or construction.
• When the buying process involves several people and you cannot tell who pays, who uses and who can block.
• When regulation, certification or technology could kill the idea, and you need to know early.

Experts complement customers. Run customer interviews for the lived problem and use experts to check the context around it. If you need an ongoing group of advisers after launch, see the customer advisory board SOP.

What do you need before you start?

• A list of your riskiest assumptions about the industry, written as statements that can be true or false: "Clinics choose software without the hospital IT team."
• A list of expert types who would know each answer.
• A one-paragraph intro about you and your research, with no pitch.
• A budget decision: free outreach through LinkedIn and associations, or a paid expert network such as GLG or AlphaSights, which charge per call and are priced for consultancies and investors.
• A notes template (below) with a column for facts and a column for opinions.

How do you run it, step by step?

1. Match assumptions to expert types

For each risky assumption, write who would know: "Rules on storing patient data" goes to a data protection consultant; "How clinics buy software" goes to a supplier sales rep or a practice-management consultant. Aim for at least two expert types per assumption so you never rely on one view.

2. Find and contact 15 to 20 experts

Search LinkedIn by job title and industry, check speaker lists from trade conferences, association boards and authors of trade articles. Send a short, specific message: who you are, what you are researching, why them, and a request for 30 minutes. Expect a third or fewer to reply.

3. Prepare questions that draw on their wide view

Ask "Across the clients you have seen…" and "What tends to go wrong when…?" questions. Ask about failed attempts: "Who tried to fix this before, and what happened?" Ask about money and decisions: "Who signs off on a purchase like this, and who can stop it?"

4. Stay clear of confidential information

Never ask a current employee of a competitor or a customer for internal numbers, plans or contracts. Say at the start that you do not want anything confidential. Former employees can share their general view of the industry, never their old employer's secrets.

5. Separate facts from opinions in your notes

Write facts (rules, prices, processes, numbers they have seen) in one column and opinions ("this will never work") in another. Opinions from experts carry weight only when they come with a reason you can check. Experts also dismiss ideas that later work, so ask for the reason every time.

6. Ask for introductions every time

End every call with: "Who are two customers or other experts I should talk to?" Experts who serve many customers are the best source of warm introductions in B2B.

7. Cross-check and update your assumption list

After each call, mark each assumption as confirmed, contradicted or unclear, with the expert who said so. When two experts disagree, add a third or ask customers directly.

Template you can copy

EXPERT OUTREACH MESSAGE
"Hi [name], I'm [your name], researching how [industry] handles [area].
Your work at [company / role] means you've seen this across many
[clients / sites]. Could I ask you 5 questions in 30 minutes?
I'm not selling anything and won't ask for anything confidential."

EXPERT INTERVIEW GUIDE (30-40 min)
1. How did you come to work with [customer group]? How many do you see?
2. Across those [clients], what problems come up again and again?
3. Walk me through how a [customer] buys a [type of product].
   Who suggests, who decides, who pays, who can block?
4. Which rules or certifications matter for a product like this?
5. Who has tried to solve [problem] before? What happened?
6. What would make you sceptical about a new solution here?
7. Which numbers would you look at to judge if this is a real business?
8. Who are two customers or experts I should speak to next?

NOTES
Expert | Type | Facts (checkable) | Opinions (with reason) |
Assumptions confirmed | Assumptions contradicted | Introductions

Worked example: software for small care homes

This is an illustrative case with invented numbers. A founder wants to sell staff-scheduling software to small residential care homes. Her four riskiest assumptions: owners choose software alone, budgets exist, no certification is needed, and nobody has a good product yet.

She contacts 18 experts and speaks to 7: two care consultants, a payroll supplier, a sector association officer, a former regional inspector, a care-home owner who sits on an association board and a sales manager from a larger software company that has since left the segment.

Results: 5 of 7 confirm owners decide alone in homes under 40 beds. Three experts contradict "no certification": scheduling must connect to staffing-ratio records that inspectors review. The former sales manager explains why her company left: homes would only pay after a successful inspection, and sales cycles took nine months. The founder gets 6 owner introductions from 7 calls, and changes her first test to an inspection-ready staffing report.

What should you measure, and when do you stop?

Stop when each risky assumption is confirmed or contradicted by two or more experts. Then go straight to customers with the sharpened questions.

Which mistakes should you avoid?

• Treating expert enthusiasm as demand. Enthusiasm costs an expert nothing. Only customers who spend money or time show that a problem is worth paying for.
• Talking to one expert. One view, even a senior one, is an anecdote.
• Asking for confidential data. It is unethical, can break their contract and puts your reputation at risk.
• Pitching your idea for approval. Ask about the industry and its history, and keep your solution for the last five minutes, if at all.
• Ignoring the "this failed before" story. A failed predecessor is the cheapest lesson you will get.

How can AI speed this up?

Use an AI assistant to turn your assumption list into an expert-type map and a first draft of questions for each type. Ask it to find trade associations, conferences and publication titles in your industry, then check each one exists. After calls, paste anonymised notes and ask it to update your assumption table with who said what. Verify every fact an AI adds with a second source, because AI tools invent regulations and industry figures with confidence. The F/MS AI co-founder helps you prepare assumption lists and question sets.

Where does this SOP fit in your validation plan?

Expert interviews run early and in parallel with customer work.

1. List assumptions and run customer interviews.
2. Check the industry context with 5 to 8 experts.
3. Use what you learn about buyers and budgets in market sizing.
4. If experts raise technical risks, run a technical feasibility assessment next.

See all 15 Customer Discovery SOPs. The Fe/male Switch startup game has first-time founders contact real experts as part of its validation tasks, with feedback on outreach messages.

FAQ

Who counts as an expert for startup validation?

Anyone with a wide view of your customers: advisers, suppliers, distributors, association leaders, regulators, journalists, community moderators and former insiders of companies in the space.

Should I pay for an expert network?

Only if free outreach fails or you need a very specific profile fast. Expert networks such as GLG and AlphaSights charge per call at rates set for consultancies, which is a lot for an early founder.

How many expert interviews do I need?

Five to eight across at least three expert types, or until every risky assumption has two independent answers.

What is the difference between an expert interview and a customer interview?

A customer interview learns about one person's lived problem and spending. An expert interview learns how the whole industry works, what failed before and who to talk to next.