The Assumption-Break PM Interview
Give a case study too big to solve, break one assumption, and watch whether they rattle off the implications.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 92%
Henrickson's compressed PM hiring loop, built to test mental agility under incomplete information rather than case-study polish. The process is deliberately short — recruiter, hiring manager, a product discussion with the CPO, and a case study. Two signals dominate: how a candidate reacts when you change one assumption mid-case, and the insightfulness of the questions they ask you, which he elicits before the product discussion rather than after.
Origin
Jeremy Henrickson's own hiring practice at Rippling, designed to select for the specific trait his environment demands: making good decisions with incomplete information in a complex decision space, which maps to the Rippling/Amazon leadership principle 'leaders are right a lot.'
Core principles
- 01Deliberately design the case study to be too complex to fully answer — the space of the problem must exceed what anyone can hold up front.
- 02The prompt should always reflect the actual business the candidate is coming into, not a generic abstraction.
- 03Change one assumption and watch: strong candidates blink twice and start rattling off the 400 downstream implications; weak ones get flummoxed.
- 04Ask 'what questions do you have for me?' before the product discussion, not at the end.
- 05Question quality is a compound signal — of genuine interest, of research done, and of depth of thinking.
- 06The best signal is when a candidate's question makes you pause and think harder about your own answer.
- 07'Right a lot' is nearly untestable in an interview and largely unteachable — you validate it over time, and by asking people who worked with them.
- 08PMs must think as business leaders: revenue and strategy questions AND detailed implication-chasing questions.
How to run it
- 1
Keep the loop short
Run a compressed process: initial contact, a conversation with the hiring manager, a product discussion with the CPO, and a case study. That is the whole thing, modulo conversations around the edges.
Pro tip A short loop is itself a filter — it signals the tempo of the company to the candidate.
- 2
Open with 'what questions do you have for me?'
Before running the product discussion, ask the candidate what questions they have for you. Do it first so the questions reveal what they arrived with rather than what your discussion just handed them.
Pro tip It does not have to be perfect. Watch both prepared questions and how well they generate questions on the fly.
Watch out Asking this at the end, as a courtesy closer, destroys the signal entirely.
- 3
Grade the questions on insight, not politeness
Evaluate whether questions show real interest, real research, and real depth. The strongest tell is when you think 'I would never have thought to ask that' and have to pause before answering.
Pro tip Look for both altitudes: business-level questions (revenue, strategy) and detail-level questions that chase implications of something you said earlier.
Watch out Expect variance. Not every question needs to be brilliant; you are looking for the presence of at least one that moves you.
- 4
Run a case study that is deliberately unsolvable up front
Build a case study grounded in your actual business whose problem space is too large for the candidate to have all the answers. The point is not the answer; the point is that the size of the space creates room to probe.
Watch out A tidy, solvable case study measures preparation, not thinking.
- 5
Break one assumption and watch
Mid-case, ask ad hoc questions and change a single assumption. Observe how the candidate reacts. This is the most indicative test of how deeply they understand a new problem and how mentally agile they are.
Pro tip The pass signal is concrete: they blink a couple of times and start rattling off the implications. The fail signal is being flummoxed.
- 6
Accept that 'right a lot' can't be interviewed for
The trait you most want — going into an ambiguous situation with incomplete information and a complex decision space, listening to everyone, making the call against dissent, and being right a year later — is very hard to test for. Approximate it by asking former colleagues whether this person was usually right, and by validating over time in the role.
Pro tip Reference questions should be literal: 'was this person usually right, and how do people think about that?'
Watch out Henrickson believes this is essentially unlearnable — you either make those calls well or you don't. Do not hire on the hope of teaching it.
In the wild
Three years before this interview, Henrickson was interviewing a candidate, Kyle Boston, who asked something to the effect of: if you have all these products sitting on an employee system of record, shouldn't you be thinking about pillars of underlying platform technology — permissioning and the like? Rippling had not yet fully formalized the concept of its platform beyond the employee system of record. Henrickson's reaction was: yes, yes we should.
→ The question — from someone with almost no context, weeks into interviewing at several companies — gave Henrickson immediate confidence in his ability to think through exactly the problems Rippling needed him for. Boston now runs Rippling's platform product organization.
Common mistakes
Saving 'any questions for me?' for the end
Asked last, it becomes a formality and the candidate's questions are contaminated by everything you just told them. Asked first, it is one of the highest-signal questions in the loop.
Designing a case study that can actually be solved
A tractable case rewards rehearsal. An intractable one creates the surface area you need to ask ad hoc questions and break assumptions — which is where the real signal lives.
Believing you can test for judgment
'Right a lot' is a peculiar skill that Henrickson says you largely can't learn and can't reliably interview for. Trying to test it directly wastes the loop; validate it through references and over time in role.
Is it for you?
Best for
Hiring managers and product leaders recruiting senior PMs into high-ambiguity, high-tempo environments where decisions must be made with incomplete information.
Not ideal for
Roles in stable, well-specified domains where execution reliability matters more than mental agility under shifting assumptions, or junior hiring where candidates have no domain base to reason from.
From the transcript
“there's a part where they they do this case study and an important part of the case study is that it's like actually too complex…”
“like ask ad-hoc questions or to like change one assumption and seeing how people react to that is really indicative of How Deeply they understand…”
“there's always a part of my interview which is which is uh maybe sounds very simple but it's just like hey what questions do you…”
“the other thing that really matters to me is like the insightfulness of the questions that people ask”
From the episode
Moving fast and navigating uncertainty
Jeremy Henrickson (Rippling, Coinbase)