Affirm PM interview: the credit-risk/UX tradeoff is the job

Affirm’s PM interview tests one thing before all else: whether you understand that credit is the product, not a constraint that lives in a compliance drawer somewhere. Every product decision at Affirm sits in a triangle formed by merchant conversion goals (approval rates up), risk team goals (loss rates down), and consumer trust goals (“no fine print”). Generic fintech prep does not prepare you for this. The candidates who get offers can hold all three simultaneously and propose something that moves one without catastrophically harming the other two.

Affirm hit $47.5B projected GMV in FY2026 (up from $36.7B in FY2025 and $20.2B in FY2023), reached its first GAAP operating profit in Q1 FY2026 at $63.7M, and reduced 30+ day delinquencies by 187 basis points in 2025. That trajectory came from specific product bets: Adaptive Checkout, AdaptAI, BoostAI, the Affirm Card. Know what those products do before you walk in.

The interview process

Five stages, rated 3/5 difficulty.

Recruiter screen (30 min). Background and motivation. Affirm recruiters will probe why BNPL and why Affirm specifically. “I find fintech interesting” is not an answer. Name the merchant-pay model and what it means for product design.

Hiring manager behavioral round (60 min). Affirm’s values (People come first, No fine print, It’s on us, Simpler is better, Push the envelope) are not decoration. The HM maps your stories directly to these. A story that ends with “we added a disclosure” does not satisfy “no fine print”: that value means the product is designed so that the terms are impossible to miss, not that the terms are technically present. Come with a story where you made complexity visible to a user who hadn’t asked to see it.

Execution round (60 min). Metrics, root cause analysis, prioritization under constraint. The critical tell: Affirm’s actual north-star metrics are GMV, revenue less transaction costs (RLTC), and loss rate. A candidate who says “I’d track NPS and monthly active users” signals they haven’t studied the domain. Know what RLTC is. Know that a feature which increases approval rate while increasing loss rate may destroy RLTC even if GMV grows.

Product sense round (60 min). This is where the three-way tension gets tested directly. See the strong/weak section below for what this looks like in practice.

Cross-functional behavioral round (60 min). How you work with risk, legal, engineering, and merchant partners. Affirm’s regulatory context is real: CFPB research shows BNPL users are statistically more likely to hold payday loans and bank overdrafts. A PM who treats compliance as a speed bump will fail here. The question is whether you can build for this population while still growing GMV.

What the product sense round is really testing

The 2026 angle: the PM job has changed shape

Affirm’s AdaptAI (launched 2025) auto-selects the optimal financing offer per transaction in milliseconds. BoostAI runs automatic A/B tests of financing offers across merchants, producing 5-15% GMV increases for participants. Adaptive Checkout drove a 26% increase in cart conversion and 22% lift in approvals for merchants who adopted it.

The PM job at Affirm in 2026 is not primarily a checkout UX job. It is a machine learning product job where the decisions are about which signals to feed AdaptAI, how to set the reward functions for BoostAI, and when to override model-driven approval decisions with human-readable policy. Feasibility is no longer the constraint: Affirm can build almost any ML-driven underwriting feature. The constraint is whether the feature serves a market willing to pay (merchants, investors) while remaining genuinely trustworthy to a population that is, statistically, financially vulnerable.

The candidate who gets the offer can articulate the job in one sentence: viable means GMV and RLTC grow; lovable means consumers feel the product is on their side rather than extracting from them; and the PM’s job is to define “optimal” for the model objective in a way that satisfies both. See proving viability, lovable, not just usable, and feasibility is free for the framing that underlies this.

Know these products before the interview

Affirm’s merchant discount rate (MDR) is approximately 6% plus $0.30 per transaction, paid by merchants. Consumers pay zero fees. This revenue model shapes every product decision: features that increase merchant approval rates have clear ROI; features that only improve consumer sentiment without touching conversion do not. The Fiserv partnership (January 2026) is bringing pay-over-time to debit card programs across financial institutions. The Affirm Card (Visa debit) drove roughly one-third of GMV growth in FY2025. These are not trivia items. They are the context required to make a product proposal sound like it came from someone who did their homework.