What should you ask about their operating history?

Five questions, and they come first because they calibrate everything that follows. Someone who has held the seat answers with a date, a number and a consequence; someone who has only advised answers with a category. The distinction matters because a recommendation is cheap for the person who never has to live inside it. You are not testing whether they are smart. You are testing whether they have ever been wrong in a way that cost them something, because that is the experience that makes the second opinion worth a five-figure quarter.

• What product did you own, at what stage, and what were you accountable for?
• What did you ship that worked, and how did you know it worked?
• What did you kill, and what did killing it cost the company?
• What is the most expensive product mistake you have personally made?
• How many companies are you working with right now?

The answers that should worry you: a career of advising with no operating seat; "it drove significant engagement" with no number attached; nothing ever killed; a most-expensive-mistake story where the mistake belonged to someone else; and a portfolio large enough that your engagement is a rounding error in their week. On the last one there is no universally right number, but there is a wrong kind of answer, which is vagueness.

What should you ask about their method?

Three questions covering evidence, prioritization and exit. Method is where the difference between an operator and a performer becomes visible, because a real method has edges: it names how many conversations before a pattern counts, what the tie-break is when two initiatives score equally, and what physically gets handed over at the end. Anyone can describe a philosophy of customer-centricity. Far fewer can tell you the number of interviews at which they stop hearing new things, which in a serious engagement lands around 60–80.

• How do you gather customer evidence, and how many conversations before you trust a pattern?
• What is your prioritization rule, and what happens when your largest customer disagrees with it?
• What does the handover look like, and what does my team still have ninety days after you leave?

The worrying answers are all versions of dependency. Evidence gathered by your team and interpreted by them. A prioritization rule that reduces to their judgment call, which means the rule leaves when they do. A handover described as a document rather than a running cadence someone on your side owns. If the honest answer to the third question is "you would keep me on", you have learned something useful about the engagement you were about to buy.

You are not testing whether they are smart. You are testing whether they have ever been wrong in a way that cost them something.

What should you ask about the engagement itself?

Two questions, and they are the ones founders most often skip because they feel adversarial. They are not. Both protect the consultant as much as you: an engagement with named success criteria is one that can be defended when a quarter goes sideways, and an engagement with a stop clause is one nobody has to stay in out of politeness. Ask them in the first call, not in the contract review, because the quality of the answer tells you whether the person has done this enough times to have been burned by its absence.

• What does success look like in numbers at 90 days, and which of those numbers can realistically move by then?
• What happens if it is not working, and what would you want the stop clause to say?

The second question has a tell. A candidate who has run real engagements will add to your stop clause rather than negotiate it down, because kill criteria make good work legible. The worrying version is enthusiasm about lagging metrics inside 90 days: anyone promising revenue or churn movement in the first quarter is either scoping something trivial or selling. What should move first is covered in the guide to measuring a product consultant in 90 days.

What does a strong set of answers look like overall?

Specific, and occasionally unflattering. The strongest candidates volunteer the engagements that did not work and can say why in a sentence that implicates their own judgment rather than the client's organization. They ask you more questions than you ask them, and the questions are about revenue, churn reasons and who currently decides what gets built, not about your tech stack. They will tell you which parts of your problem they are not the right person for. Product management consulting for B2B SaaS is the category; a fractional CPO is the engagement model inside it that carries accountability for the outcome, and a candidate who can explain which of the two your situation calls for is demonstrating the judgment you are buying.

Sivan Kadosh, a fractional Chief Product Officer for B2B SaaS companies between $2M and $15M ARR, works on a $8,000 monthly retainer covering 20–25 hours. The process end of this is in how to hire a product management consultant, the types and costs are in should you hire a product consultant, and the selection criteria are in how to choose a product management consultant.

Ten questions, in order: five on what they have owned, three on how they work, two on how the engagement is judged and ended. Specificity is the signal throughout, and the stop-clause question is the single most informative of the ten. Ask all of them in the first call, before anyone has written a proposal.