Meta Product Manager Interviews: The New Product Sense with AI round
Meta recently added a fourth interview to its final PM loop. It’s one of the larger changes in the last couple of years (we don’t count the renaming of Product Execution to Analytical Thinking a real change, since the format, rubric, expectations etc didn’t really change). This new round is called "Product Sense with AI". It seems to be showing up primarily for Meta PM roles at the IC6 (senior+ individual contributors) and M1 / M2 levels (i.e. people manager) level, particularly for roles which will double down on AI (either from a technical/infra POV - or from a user needs POV).
Everything you need to win this interview
Try Free AI Interview
What to expect in the Meta "Product Sense with AI" interview
You'll get a product sense case, but you are expected to use AI in the interview when answering the different components you go through. The format itself is similar; so you’d still be expected to talk through the product motivation, the customer segmentation / targeting, narrowing down on a solution etc; except you will also be assessed on how you leverage AI tooling to answer (or improve the quality of your answer) for each of these. And it’s really important to use this the right way upfront, since whatever you start with (eg. the customer you choose after during segmentation), has a major impact on the rest of your interview).
if you mess one part up earlier on in the stage, it has cascading effects and you really want to be conscious of that.
What good AI collaboration looks like in a Product Sense interview
For clarity - Meta isn't testing whether you can write clever prompts. Much more importantly, they're testing whether you can think with AI. Can you guide AI with the right degree of prompting and nudging? Do you critique AI suggestions instead of accepting them blindly? Can you explain why an AI-generated idea makes sense or doesn't?
The worst thing we suspect you can do, is to use AI to outsource your thinking. It’s one thing to ask AI to sanity check your idea, or to think of edge cases where it might fall through the cracks; it’s quite another to ask “what would the next best thing be”, even if it’s a sophisticated prompt asking the AI to do that. Candidates who do the latter are going to knock themselves out of the running. The best candidates treat AI as a thinking partner, they push back on its suggestions, ask it clarifying questions, and synthesize its outputs with their own product judgment.
Vikash Rungta, a former Meta Product Lead (working within their LLAMA ecosystem), actually talks through the key components which have changed.
How to prepare for the Product Sense with AI interview
If you're interviewing for a meta product manager position, start by answering product sense questions yourself first. After you've worked through the entire problem, mission, segmentation, pain points, solutions, then ask AI to tackle the same question, not in its entirety, but within the context of each individual section. Compare your approach to AI's approach and identify specific gaps. Was your segmentation too narrow? Did you forget to consider technical feasibility? This comparison method builds judgment, not dependence. (Incidentally, you can do a pretty close simulation of this using Prepfully’s Meta AI interviewer; just make sure to choose Product Sense)
Get familiar with AI product constraints. AI outputs are probabilistic, not deterministic. They have variable costs (inference, compute, latency) that traditional software doesn't. The product sense with ai format will expect you to know the limitations, know the potential, and incorporate these within your answer - in some cases you’ll need to guardrail your prompts, and in others you’ll need to encourage the AI to think beyond. A really deep familiarity with using AI in the context of exploring new ideas (which Product Sense interviews typically expect of you) is therefore not optional.
The best way to prepare is to practice with someone who's actually conducted these interviews.
Prepfully has Meta PM coaches who can simulate the Product Sense with AI format.
→ Schedule now
tl;dr - 4 tips for acing the Meta Product Sense with AI interview
- Practice the compare-and-contrast method: Answer product sense questions yourself first, then use AI to compare approaches and spot your gaps. This approach is critical for any meta product manager candidate.
- Understand AI system constraints: Know how non-deterministic outputs, inference costs, and latency affect product decisions.
- Use multiple tools before the interview: Familiarity and knowledge of different tools’ (Meta AI, Claude, Perplexity) strengths and weaknesses matters, don't walk in having only used ChatGPT.
- Critique everything AI suggests: As a meta product manager, the interview tests your judgment, not simply your ability to prompt. Always explain why an AI suggestion does or doesn't work.