Every STR listing eventually produces a number: "projected $85K/year!" Your job is not to repeat that number. Your job is to know whether it survives scrutiny — because your client's lender, spreadsheet, and future self all will scrutinize it.
STR revenue estimation works like a CMA, but on performance data: find 5–10 actually comparable active listings — same bedroom count, sleeps, amenity tier, and micro-location — and look at their trailing-12-month revenue, not their best month. Data tools estimate this; calibrated common sense filters it. A comp with a hot tub and lake view is not a comp for a property with neither.
Annual revenue ≈ average daily rate × nights booked. The dishonesty usually hides in occupancy: projecting 75% in a market that averages 55%, or using peak-week ADR as the annual average. Recalculate every projection with market-median occupancy and blended ADR. If the deal only works at aspirational numbers, it doesn't work.
When a seller reports revenue, ask for the platform earnings reports or Schedule E — not a summary spreadsheet. Then ask the harder questions: Was it owner-boosted (friends' bookings, owner blocks reducing supply)? Is the permit that produced that revenue transferable? Was last year an anomaly (event year, eclipse, championship)?
Red flags: projections based on the single best comp in the market; "potential" revenue for an unpermitted property; pro formas with no cleaning, management, or platform fees; and any projection the presenter can't decompose into ADR and occupancy on request.
See a full worked example in the STR Deal Analysis Case Study, and pressure-test numbers with the Income Estimator.
ACTION: Take one active listing with a revenue claim and decompose it: what ADR and occupancy does that number imply? Compare against the market median. Practice saying the verdict out loud.
Next lesson: the inspection items that only matter for STRs — septic math, hot tub electrical, and the insurance call that must happen before closing.
Gross revenue before expenses. Sanity-check against real nearby listings — projections are estimates, not promises.