Mercari Japan Underpriced Deals — Live vs Sold Comps
Pricing
from $20.00 / 1,000 keyword scanned for deals
Mercari Japan Underpriced Deals — Live vs Sold Comps
Which listings on Mercari Japan are cheap right now? Builds the sold-comp median for your keyword, then scores every live listing against it and returns the ones below your discount threshold, with the discount distribution and the sold-comp basis. For resellers and proxy buyers. From $0.02.
Pricing
from $20.00 / 1,000 keyword scanned for deals
Rating
0.0
(0)
Developer
h ichi
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
4 days ago
Last modified
Categories
Share
"What is it worth?" is a solved question — plenty of Actors return Mercari price statistics. The question that makes money is "what is mispriced right now?", and answering it needs both sides in the same call: the sold-comp median and the live book, scored against each other.
How the score is built
- Sample sold (
sold_out) listings for the keyword → the basis: median and quartiles. - Sample live (
on_sale) listings. discountPct = (1 - price / soldMedian) x 100for each live listing.- Return every listing at or below your threshold, plus the distribution.
Two guards, because a basis can lie in two ways
Too few comps. A median built on three sold listings is noise, and a "70% discount" against noise is a bad trade. minComps (default 8) refuses to score such a keyword.
Too many kinds of thing. This one we found by running it. Search Mercari for Nikon Z6 and you get the camera and its batteries, cases and manuals: sold comps from ¥375 to ¥150,000, median ¥7,240 — an accessory median. Scored against that, a ¥490 product catalogue booklet came back as the best deal at "93.2% off". Arithmetically perfect, completely fabricated.
So the basis must also be coherent. maxCompSpread (default 4.0) compares the p75/p25 ratio of the sold comps; a keyword that mixes price tiers is refused with the spread quoted and a pointer to the fix:
priceMinJpy / priceMaxJpy bound both sides of the comparison, so the camera is scored against camera comps.
In both cases the Actor returns basisUsable: false with basisUnusableReason, and is not charged. The basis is always reported next to the deals (soldMedianJpy, soldCompCount, compSpread, quartiles) so you can see what the discount was measured against.
Output
Per keyword, one deal_summary:
| field | meaning |
|---|---|
soldMedianJpy, soldPriceJpy | the basis and its quartiles |
basisUsable / basisUnusableReason | whether the basis held, and why not |
compSpread | p75/p25 of the comps — how mixed the keyword is |
soldCompCount / liveListingCount | sample sizes on both sides |
dealCount | live listings at or below the threshold |
dealRatePct | share of the live book that is underpriced |
discountPctStats | quartiles of the discount across deals |
bestDeal | the single largest discount, with its URL |
Plus, optionally, each deal as its own record.
Pricing
From $0.02 per keyword scanned, $0.002 per individual deal. A keyword with an unusable basis is never charged.
Read-only public data, no login, throttled requests. Nothing is stored between runs.
