What to expect, what they’re really testing, and what a strong answer looks like — scored.
Trust and safety, marketplace design, host and guest experience balance, and community. Airbnb PMs must think carefully about how product changes affect both hosts and guests — and how to build trust in a marketplace where strangers share homes.
The question below was asked by Airbnb interviewers. The answer is graded on the five dimensions real PM interviewers use: structure, specificity, reasoning, decision quality, and delivery.
“How would you reduce fraudulent listings on Airbnb?”
First, I want to scope the problem: fraudulent listings fall into two categories — fake listings (the property doesn't exist or isn't available) and misrepresented listings (the property exists but the photos or description are materially false). These have different detection mechanisms and different impact on trust.
I'd focus on misrepresented listings first because they're higher volume and harder to detect automatically — fake listings are easier to catch through address validation and booking behavior patterns.
For misrepresentation, the core problem is that hosts self-report everything: photos, amenities, location accuracy. The verification gap is widest at listing creation and during re-listings after long gaps.
Three interventions in priority order:
1. AI photo verification at listing creation: compare host-uploaded photos against Google Street View and satellite imagery for the listed address. Flag listings where the exterior doesn't match. This catches the most egregious cases and is already technically feasible with existing models.
2. Guest-verified amenity reporting: after checkout, prompt guests to confirm which listed amenities were present (pool, WiFi, kitchen equipment). Build a reliability score per amenity per listing. Surface this on the listing page. This turns guests into a verification layer.
3. Re-listing audit trigger: any listing inactive for 180+ days that is re-activated requires photo re-upload and admin review. Properties change significantly over time and stale photos are a leading cause of misrepresentation disputes.
Success metric: dispute rate per completed booking (primary). I'd track this by listing verification tier to measure whether verified listings have lower dispute rates.
Distinguishes fake vs. misrepresented listings, focuses on higher-impact category, and presents interventions in priority order.
Names specific mechanisms (Google Street View comparison, guest amenity confirmation, 180-day re-listing rule) rather than generic 'verify listings.'
Explains why misrepresentation is higher-volume and harder to automate; the guest-verification insight is non-obvious.
Prioritizes three interventions clearly; defines success metric.
Tight and well-organized; numbered list makes the priority order easy to follow.
The answer earns high marks for immediately distinguishing two types of fraud and explaining why misrepresentation is the harder, higher-impact problem. The guest-as-verification-layer idea is creative and non-obvious. The weakest part is the success metric — dispute rate per booking is correct but the answer doesn't name a baseline or target improvement, making it hard for the interviewer to judge whether this is a meaningful improvement.
Quantify the success metric: name a baseline dispute rate and the target reduction you'd expect from each intervention.
Paste any Airbnb PM interview question and your answer. Get scored on the same five dimensions — instantly, free, no signup.
Grade my answer free →First grade is free. No account needed.