How the Scout appraises a find
Same item. The seller had no idea. Neither did the people who scrolled past it. The model reads the spec out of the photos and one lazy sentence of description, then prices it against sales that actually completed — and publishes the sample size so you can check its work.
01
People sell $6,000 bikes with eight words of description. The model reads the spec out of the photos and the copy — a carbon wheelset versus alloy is a four-figure difference the seller often never mentions.
02
Asking prices are fantasy. We compare against completed sales, matched on spec instead of title text. Anyone can string-match a headline. That's why everyone else's deals aren't deals.
03
One confident number is how you hide a thin comp set. We publish the range and the sample size on every listing, so you always know whether we are certain or guessing.
04
A price far under every comp usually means the listing is wrong, not that the seller is generous. When we see that shape, we say so on the row — we don't quietly drop it to keep the board looking rich.
Pulled live from the board — this is the current best-evidenced listing, not a worked example.
We don’t have this number yet, so we’re not going to print one. Tracking what our finds actually sell for — and publishing the median error against it — is the next thing we build. When it exists it goes here, wrong answers included. Until then, judge us on the sample size and spread we publish on every listing.