Japan Sell-Through Rate — Sold vs Listed on Mercari & Yahoo

Before "what is it worth", the question is "does it sell". One call per keyword measures Japanese resale demand: what share of the listings have sold, how fast they sold, and what separates the prices that sold from the prices still sitting there.

Market	What it can measure	Basis
Mercari Japan (メルカリ)	Sell-through rate — sold-out vs on-sale listings, days to sell, listing age, sell-through per price band	stock_ratio
Yahoo! Auctions Japan (ヤフオク)	Sales per day, days since the last sale, realized prices, bid competition, instant-buy take rate	sold_flow
PayPay Flea Market (PayPayフリマ)	Competing supply — live listings and what they ask	supply_only

The one thing to read first: demandBasis

The three markets do not publish the same kind of evidence, so their headline numbers are not interchangeable:

• stock_ratio (Mercari) — a share of the current listing stock. Unit: a fraction.
• sold_flow (Yahoo! Auctions) — sales per day. Unit: 1/day.
• supply_only (PayPay) — no demand measurement at all.

A share and a rate-per-day cannot be averaged, so no field ever blends them. Every key in demand names the market it came from.

Yahoo! Auctions cannot produce a sell-through rate, and this Actor does not pretend it can. Yahoo's closed-auction index (closedsearch, 落札相場) serves auctions that sold — measured 2026-08-07 over 200 auctions across four unrelated keywords, the minimum bid count served was 1. Ended-without-a-bid auctions are simply not indexed, so any "Yahoo sell-through rate" is 1.0 by construction. sellThroughRate is therefore null there, with the reason attached, and soldPerDay is used instead.

What you get

demand.mercariSellThroughRate      0.5515      sold / (sold + on sale), from Mercari's own hit counts
demand.mercariRateBound            "exact"     exact | lower_bound | upper_bound | unknown
demand.mercariMedianDaysToSell     6           how fast the sold half moved
demand.mercariMedianListingAgeDays 6           how long the unsold half has been sitting
demand.mercariSoldMedianJpy        14250       the price that actually sells
demand.mercariOnSaleMedianJpy      8350        what live sellers are asking
demand.mercariAskPremiumPct        -41.4       ask vs realized
demand.yahooSoldPerDay             5.52        completed auctions per day
demand.yahooDaysSinceLastSale      0.76        staleness: nothing sold for N days = no demand
demand.yahooSoldAuctionsIndexed    1102        how many completed sales Yahoo has on file
demand.paypayLiveListings          589         supply you would compete with on a second platform

Plus, per market, the full distributions (min / p25 / median / p75 / max / average) for sold prices, asking prices, days to sell, listing age, bid counts and likes — and the price-band table below.

Sell-through by price band

Does the cheap end clear faster, or is it clogged with parts nobody wants? Measured for 象印 圧力IH炊飯器 on 2026-08-07:

Band	Sold (sampled)	On sale (sampled)	Sell-through
≤ ¥6,000	23	32	46.9%
¥6,000–¥11,000	18	30	42.5%
¥11,000–¥19,000	31	22	63.4%
> ¥19,000	28	16	68.3%

Here the expensive bands clear better. Below ¥6,000 the keyword catches spare parts (内ブタ, 内鍋, 雑穀フィルター), damaged units ("内釜剥がれあり") and older non-pressure models — not the machine most buyers came for. That is also why askPremiumPct came out negative: the sold pool and the on-sale pool hold different products. The band table is the fix for that composition effect, and it is why this Actor has no price filter (see Notes & limits).

Pricing — from $0.02 per keyword, no subscription

Event	Price	When
Keyword analyzed	$0.02	Per keyword, whatever the number of markets
Individual listing	$0.002	Only if you enable Include individual listings

A default run (1 keyword, 3 markets, summary only) costs $0.02 and took 8.8 seconds measured. Apify's $5 monthly free credit covers about 250 keywords at that rate. No subscription.

A keyword is only charged when at least one market actually measured demand — a sell-through rate, a sales-per-day figure, or a sold-vs-listed price gap. A run that could only count live listings measured supply, not demand, and is free.

Input

Field	Example	Notes
keywords	["象印 圧力IH炊飯器"]	Japanese matches far more listings. Narrow keywords give sharper rates. $0.02 each
markets	["mercari","yahoo_auction","paypay_flea"]	All three by default. The price per keyword does not change
maxItemsPerMarket	100	30–300 sampled per side. Does not affect the headline rate or the listing counts
includeIndividualItems	false	Enable to also get every sampled listing, +$0.002 each
convertToUsd	true	Adds USD stats at the current rate (one FX lookup per run)

Output example (type: "sell_through_summary", abridged, measured 2026-08-07)

{
  "type": "sell_through_summary",
  "keyword": "象印 圧力IH炊飯器",
  "demand": {
    "mercariSellThroughRate": 0.5515, "mercariSellThroughPct": 55.1,
    "mercariRateBound": "exact",
    "mercariSoldListings": 2274, "mercariOnSaleListings": 1849,
    "mercariMedianDaysToSell": 6, "mercariMedianListingAgeDays": 6,
    "mercariSoldMedianJpy": 14250, "mercariOnSaleMedianJpy": 8350,
    "mercariAskPremiumPct": -41.4,
    "yahooSoldPerDay": 5.52, "yahooSoldAuctionsIndexed": 1102,
    "yahooDaysSinceLastSale": 0.76, "yahooSoldMedianJpy": 6708,
    "yahooMedianBids": 6,
    "paypayLiveListings": 589, "paypayAskMedianJpy": 12675,
    "note": "mercariSellThroughRate is a share of listing stock and yahooSoldPerDay is a flow per day — different units, never average them."
  },
  "markets": {
    "mercari": {
      "demandBasis": "stock_ratio",
      "soldListings": 2274, "onSaleListings": 1849,
      "sellThroughRate": 0.5515, "rateBound": "exact",
      "countsCapped": {"sold": false, "onSale": false},
      "soldPriceJpy":   {"min": 2900, "p25": 7375, "median": 14250, "p75": 20500, "max": 67999, "average": 16757},
      "onSalePriceJpy": {"min": 500,  "p25": 5500, "median": 8350,  "p75": 14999, "max": 97900, "average": 11774},
      "priceGap": {"soldMedianJpy": 14250, "onSaleMedianJpy": 8350, "askPremiumJpy": -5900, "askPremiumPct": -41.4,
                   "sellingBandJpy": {"p25": 7375, "median": 14250, "p75": 20500},
                   "unsoldSide": "open_ask"},
      "speed": {"medianDaysToSell": 6, "soldWithinDays": 7, "soldWithinDaysPct": 57.0,
                "daysToSell": {"p25": 3, "median": 6, "p75": 14, "max": 262},
                "medianListingAgeDays": 6, "olderThan30DaysPct": 17.0},
      "priceBands": [
        {"label": "≤ ¥6,000", "soldSampled": 23, "onSaleSampled": 32, "sellThroughPct": 46.9, "basis": "sampled_scaled"},
        {"label": "> ¥19,000", "soldSampled": 28, "onSaleSampled": 16, "sellThroughPct": 68.3, "basis": "sampled_scaled"}
      ]
    },
    "yahoo_auction": {
      "demandBasis": "sold_flow",
      "sellThroughRate": null,
      "sellThroughRateNote": "Yahoo! Auctions' closedsearch index contains auctions that sold …",
      "soldAuctionsIndexed": 1102, "sampledAuctions": 100,
      "soldPerDay": 5.52, "sampleWindowDays": 17.94,
      "newestSoldAt": "2026-08-06T21:55:21+09:00", "daysSinceLastSale": 0.76,
      "soldPriceJpy": {"p25": 3675, "median": 6708, "p75": 18150, "max": 93000},
      "bidCount": {"p25": 1, "median": 6, "p75": 15, "max": 50},
      "instantBuy": {"auctionsWithInstantBuyPrice": 15, "closedAtInstantBuyPct": 26.7,
                     "medianPriceToInstantBuyPct": 79.7}
    },
    "paypay_flea": {
      "demandBasis": "supply_only", "sellThroughRate": null,
      "liveListings": 589, "askPriceJpy": {"median": 12675}, "listingsWithLikesPct": 82.0
    }
  },
  "marketsWithDemandSignal": ["mercari", "yahoo_auction"],
  "definitions": { "sellThroughRate": "…", "soldPerDay": "…", "askPremiumPct": "…" },
  "checkedAt": "2026-08-07T07:03:40+00:00"
}

With includeIndividualItems: true you additionally get one record per sampled listing, tagged with market and side (sold / listed).

Use cases

• Sourcing decisions — a 70% sell-through with a 6-day median is a product you can buy; a 20% sell-through with 200-day-old listings is dead stock whatever the price says
• Pricing to actually sell — priceGap.sellingBandJpy is the p25–p75 of what changed hands, and priceBands shows where the rate falls off
• Deciding where to list — Mercari's rate and Yahoo's soldPerDay describe two different sales mechanics; paypayLiveListings is the competition on a third
• Inventory triage — schedule the Actor over your SKU list and watch sellThroughRate and daysSinceLastSale move
• Demand research — how does sell-through vary by price band and category across the Japanese resale market

Notes & limits — how the statistics are defined

These are sample-based statistics, not a census. Read this before quoting a number.

• sellThroughRate is a stock ratio, not a conversion rate. It is sold-out listings / (sold-out + on-sale listings) from Mercari's own hit counts for the keyword — a snapshot of the listing population, not "N% of listings sell within X days". It is not comparable with Yahoo's soldPerDay.
• Mercari caps its hit count at 15,000, so a rate built from a capped count is a bound and rateBound says which: lower_bound (sold side capped), upper_bound (on-sale side capped), or unknown — in which case the rate is null, because two capped counts force exactly 0.5 by construction. Measured: nintendo switch returns 15000/15000 and gets no rate; ライカ M6 returns 396/297 and gets an exact 57.1%.
• soldPerDay is measured over the sampled window, so always read sampleWindowDays beside it. On a huge keyword 30 samples can span 0.03 days and the extrapolated rate is instantaneous, not a daily average. Raise maxItemsPerMarket to widen the window.
• Price, speed and bid statistics come from a sample of up to maxItemsPerMarket listings per side, taken in each site's default result order. The listing counts and the sell-through rate do not.
• priceBands is an estimate. Each band weights its sampled sold and on-sale counts back up by that side's total listing count, because the two sides are sampled at different rates. soldSampled / onSaleSampled are published so a band resting on three listings is visible as such. Mercari's own hit count could not be used here: measured 2026-08-07, adding a price ceiling makes it repeatably non-monotone (the sold side answered 1,604 for ≤¥3,000 but 1,101 for ≤¥6,000), so a ladder of ceilings cannot be differenced into band counts.
• This Actor deliberately has no price filter. A client-side price window would filter the sample but not the site's hit counts, silently breaking the rate it is supposed to support. The price-band table answers the same question without corrupting anything.
• Mercari's daysToSell is updated - created, which also moves on price edits, so it skews low — a liquidity signal, not an exact time-to-sale. The age of a live listing uses created and the wall clock instead, so an edit cannot make a stale listing look fresh.
• A negative askPremiumPct usually means product mix, not a bargain: the on-sale pool holds accessories and parts that the sold pool does not. Use priceBands.
• Yahoo's instantBuy excludes fixed-price listings, where price == buyNowPrice by definition and a "take rate" would be meaningless.
• A market that fails does not fail the run: it comes back as available: false with an error, and the other markets still report. Only a keyword where no market measured demand fails the run.
• Read-only, no login, no browser. Requests are throttled to one per 1.5 s, and Yahoo! Auctions and PayPay Flea Market are queried sequentially because they share one backend.
• No personal data is collected — sellers are excluded from every record.

---

日本語ガイド

キーワード1件につき、日本の中古市場の「売れ行き(消化率)」を1レコードで返す Actor です。「いくらか」ではなく「そもそも売れているのか」に答えます。

• メルカリ: sellThroughRate = 売り切れ件数 /(売り切れ+販売中)。メルカリ自身の検索ヒット件数から算出します。加えて売れるまでの日数、売れ残りの掲載日数、価格帯別の消化率を返します。
• ヤフオク: 落札率は返しません。ヤフオクの「落札相場」は落札済みのみを収録しており(2026-08-07 実測: 4キーワード200件すべて入札1件以上)、そこから消化率を出すと必ず 1.0 になるためです。代わりに1日あたりの落札件数・最終落札からの経過日数・落札価格分布・入札数・即決価格の到達率を返します。
• PayPayフリマ: 売却済みを公開していないため供給(出品件数と提示価格)のみです。
• 数値の性質が市場ごとに違います(demandBasis)。メルカリは「在庫に対する割合」、ヤフオクは「1日あたりの件数」で単位が異なるため、平均してはいけません。全フィールドが市場名を含む形で分かれています。
• メルカリのヒット件数は 15,000 で頭打ちになります。両側が頭打ちだと比率が必ず 0.5 になるので、その場合は rateBound: "unknown" として数値を出しません(実測: nintendo switch は 15000/15000 で非公開、ライカ M6 は 396/297 で 57.1%)。
• 料金: キーワード1件 $0.02(市場を何個選んでも同額)、個別明細は $0.002/件(既定 OFF なので既定実行は $0.02 固定)。どの市場も需要を測れなかったキーワードには課金しません。サブスクリプション不要です。
• 注意: 価格・日数・入札数の統計は各サイトの既定並び順から最大 maxItemsPerMarket 件の標本です(全数ではありません)。消化率と件数のみサイトのヒット件数由来です。価格帯別の消化率は標本を母数で重み付けした推定値です(メルカリのヒット件数は価格上限を付けると単調でなくなることを実測したため、件数の差分では計算できません)。ブラウザ不使用・ログイン不要・1.5秒に1リクエスト、ヤフオクと PayPayフリマは同一基盤のため逐次取得しています。個人情報は取得しません。