eBay Product Research: Sold Prices & Sell-Through Comps

Run eBay product research on a list of product phrases and get one decision-ready row each: the median sold price with its p25–p75 band, the sell-through rate, how many competing listings are live right now, and how fast the product moves per day.

Built for resellers, dropshippers, liquidation buyers, and category managers who decide what to buy and what to charge — not for people who want 10,000 raw listing rows to pivot by hand. One search in, one row out, $0.05 per report at the volume tier.

What does the eBay comps actor do?

You give it product phrases (or full eBay search URLs). For each one it reads the sold listings inside a 7-, 30-, or 90-day window and the active listings competing for the same search, then answers the three questions behind a buying decision: what does it sell for, how often does a listing actually sell, and how fast. Prices come back as a median plus a p25–p75 band, so you see the price the market pays and the spread around it, split per item condition.

It is an aggregator, not a listing dump: it returns one row per product, not a page of offers.

Why use this eBay sell-through calculator

• Median sold price, not asking price — p25/p75 band from realized sales inside your window.
• Sell-through rate = sold ÷ active listings, the demand-vs-supply ratio eBay does not show.
• Counts you can trust — de-duplicated by item ID, and every count carries an exactness flag.
• Sales pace and trend — sold per day, plus the percent change between the newer and older half of the window, so a flat median does not hide a collapsing category.
• Per-condition breakdown — New, Pre-Owned and Refurbished each get their own count and median.
• Optional Amazon spread — when the product publishes a UPC, the row adds the Amazon price and the percentage gap against your eBay median.
• Repeat searches are free for 24 hours — the same search, window and filters is re-delivered from cache and not billed as a report.

How to check eBay sold prices and sell-through

1. Open the Input tab and paste your product phrases into Product searches, one per line (a full eBay search URL works too, and overrides the filters below it).
2. Pick the eBay marketplace — 14 sites, from EBAY_US to EBAY_GB and EBAY_DE. All prices come back in that marketplace's currency.
3. Pick the sales window: 7, 30, or 90 days. 30 days is the usual restocking window.
4. Optionally narrow by item condition and a minimum / maximum price.
5. Leave Add Amazon price spread on if you also want the Amazon comparison for products that publish a UPC.
6. Click Start and read the Overview view of the dataset, or pull the rows through the API.

Minimal input:

{
  "queries": ["logitech mx master 3s", "sony wh-1000xm4"],
  "ebaySite": "EBAY_US",
  "windowDays": "30",
  "enableAmazonSpread": true
}

What data you get per product

One row per search phrase. A real row looks like this:

{
  "query": "logitech mx master 3s",
  "ebaySite": "EBAY_US",
  "windowDays": 30,
  "windowStart": "2026-07-05",
  "windowEnd": "2026-08-03",
  "currency": "USD",
  "soldCount": 459,
  "soldCountExact": true,
  "activeCount": 853,
  "activeCountExact": true,
  "sellThroughRate": 0.5381,
  "soldPriceMedian": 60,
  "soldPriceP25": 40,
  "soldPriceP75": 79.99,
  "soldVelocityPerDay": 15.3,
  "velocityTrendPct": 12.4,
  "conditionBreakdown": [
    { "condition": "Pre-Owned", "soldCount": 198, "medianPrice": 52.5 },
    { "condition": "Brand New", "soldCount": 121, "medianPrice": 79.99 },
    { "condition": "Certified - Refurbished", "soldCount": 74, "medianPrice": 63.99 }
  ],
  "amazonUpc": "097855161581",
  "amazonAsin": "B09HM94VDS",
  "amazonPrice": 99.99,
  "amazonSpreadPct": 66.65,
  "amazonMatchMethod": "upc",
  "pagesFetchedSold": 4,
  "pagesFetchedActive": 6,
  "source": "primary",
  "status": "ok",
  "notice": null,
  "fetchedAt": "2026-08-03T12:46:54.276Z"
}

Field	Type	What it tells you
query, ebaySite, windowDays	string / string / int	Input echo — what this row answers
windowStart, windowEnd	date	The closed window that was counted, exactly windowDays long
currency	string	Currency of every price in the row
soldCount, activeCount	int	Sold in the window and live competitors now, de-duplicated by item ID
soldCountExact, activeCountExact	bool	false = the count is a lower bound, see notice
sellThroughRate	float	soldCount / activeCount, 0.54 = 54%; null when there is no denominator
soldPriceMedian, soldPriceP25, soldPriceP75	float	Realized price and its accepted band
soldVelocityPerDay	float	Average sold listings per day in the window
velocityTrendPct	float|null	Pace of the newer half vs the older half of the window
conditionBreakdown[]	object[]	{condition, soldCount, medianPrice} per condition
amazonUpc, amazonAsin, amazonPrice, amazonSpreadPct, amazonMatchMethod	mixed	Amazon comparison, UPC join only
pagesFetchedSold, pagesFetchedActive	int	How deep the review went behind the numbers
source	string	primary or fallback for the active-listing page
status, notice	string	ok / not_found / too_broad, plus a plain-English caveat
fetchedAt	datetime	When the row was measured

Two dataset views ship with the actor: Overview (the six decision columns) and Needs a narrower query, which isolates rows whose counts hit the review depth limit.

How sell-through, percentiles, and trend are calculated

Sell-through is soldCount / activeCount on the same phrase with the same filters, so it is comparable across your list. The median and both percentiles use one interpolation method on realized prices only — asking prices of live listings never enter them. velocityTrendPct compares the sales pace (sales ÷ days) of the newer half of the window against the older half, so a 30-day window tells you whether the last 15 days ran hotter than the first 15. When a count cannot be completed, the row says so in notice and flips the matching *Exact flag to false instead of publishing a number that reads as final.

Input parameters

Parameter	Type	Default	Notes
queries	string[]	—	1–100 phrases or eBay search URLs; one row per entry
ebaySite	select	EBAY_US	14 marketplaces; sets the currency of the row
windowDays	select	30	7, 30, or 90 days
condition	select	Any	New / Refurbished / Used
priceMin, priceMax	int	empty	Bounds in the marketplace currency
enableAmazonSpread	bool	true	Adds the Amazon columns when a UPC exists

The full descriptions live on the Input tab, and the same schema drives the Apify API and the MCP server at mcp.apify.com, so an AI agent can call this actor without a wrapper.

How much does eBay product research cost?

Pay per event — you pay for delivered reports, not for compute time.

Event	What triggers it	Price
actor-start	Once per run that delivers at least one report	$0.005
product-report	Each delivered row, including not_found answers	$0.08 → $0.05

The report price is a ladder across plan tiers: $0.08 on Free, then $0.07 and $0.06, and $0.05 from the volume tier up. A run with one product costs $0.085 minimum. Cache hits and rows that could not be measured are never charged.

eBay comps actor vs per-listing scrapers and manual research

	This actor	Per-listing eBay scrapers	Manual eBay search
Output	One aggregated row per product	Hundreds of raw listing rows	A page you read yourself
Sell-through rate	Yes — sold and active in one row	No, sold listings only	No
Price band	Median + p25/p75 + per condition	You compute it	Eyeballed
Cost per product	$0.05 per report	Billed per listing, hundreds per product	Free, minutes per product
100 products	One run	One run per product	Hours

eBay comps actor vs Terapeak

Terapeak is eBay's own product-research tool inside Seller Hub, and for a single product it is excellent — first-party data with up to 3 years of history, free with an eBay seller account. This actor is not a Terapeak clone; it covers the part Terapeak does not:

	This actor	Terapeak Product Research
Access	Public API — no eBay account needed	Requires an eBay seller account, browser UI
Bulk	Up to 100 products per run	One search at a time
Automation	Schedulable runs, JSON/CSV dataset, API	Manual searches, manual export
History window	7 / 30 / 90 days	Up to 3 years
Data source	Public sold + active listings (see limits below)	eBay first-party sales data
Sell-through	Sold vs active in the same row	Available
Cross-market check	Optional Amazon price spread (UPC join)	eBay only
Price	$0.05–0.08 per product report	Free with a seller account

Rule of thumb: researching one product deeply and you already sell on eBay — open Terapeak, it is free and has longer history. Screening a list of candidates, feeding a repricer or a sourcing pipeline, or working without a seller account — that is what this actor is for. Terapeak has no public API to plug into a workflow; this actor is the automatable route to the same class of answer.

How the numbers stay honest

• De-duplication by item ID, within a page and across pages — eBay repeats listings between result pages, and an undeduplicated count inflates demand by double digits.
• Every listing is date-checked against windowStart/windowEnd; sold pages mix in results far older than your window.
• Depth limits are disclosed, never silent: *Exact: false plus a notice, surfaced in the "Needs a narrower query" view.
• No invented sell-through — unmeasured sold data leaves every sold-side field null.

What this actor does not do

• It does not measure demand outside eBay. eBay is a proxy for demand, not your buyer's channel — a product that sells well on eBay can behave differently on Amazon, Shopify, or in retail, and the reverse is just as true.
• It does not fill the Amazon columns for every product. The match is a strict UPC join, and a usable UPC is published for roughly 30–40% of products (games, fashion, and spare parts are usually not among them). The eBay core columns are always filled.
• It does not match products by title — a title match returns the wrong product often enough to poison a buying decision, so it is not offered at any confidence level.
• It does not return individual listings, seller names, or buyer data. It is an aggregate.
• It does not track products over time on its own. It is stateless — schedule it to build history.
• It does not guarantee exact counts on very broad phrases; those come back flagged as too_broad with a lower bound.

FAQ

How do I check the real sold price of an item on eBay?

Enter the product phrase and a window. The row returns soldPriceMedian for realized sales inside that window, plus soldPriceP25 and soldPriceP75 so you see the spread rather than one number.

What is a good sell-through rate on eBay?

Resellers commonly treat 50%+ over 30 days as healthy demand and under 20% as a slow mover. This actor gives you sellThroughRate per product, so you compare candidates on the same scale.

Why is sellThroughRate sometimes null?

Because there was no denominator — no active listings matched, or sold data could not be measured. A null is an honest "not measured"; it is never published as a 0% sell-through.

Is this a Terapeak alternative?

For bulk and automated research, yes. Terapeak is free with an eBay seller account and has longer history for a single product, but it is a manual browser tool with no public API. This actor answers the same question — realized prices and sell-through — for up to 100 products per run, on a schedule, as a dataset, with no eBay account. See the comparison table above.

How many products can I check in one run?

Up to 100 phrases per run, one billed report each. Re-running the same phrase within 24 hours re-delivers the cached report and is not billed.

Does it work on eBay UK, Germany, or Australia?

Yes — 14 marketplaces, selected with the ebaySite field. Prices are returned in that marketplace's currency, so do not mix currencies when you compare rows across sites.

Related actors

• Amazon Listing Optimizer — when the Amazon spread says a product is worth listing there, this audits the listing that has to sell it.

See the doesaiknow publisher page for the rest of the e-commerce research suite.

Legal & support

• This actor returns aggregated public marketplace statistics — counts, medians, percentiles. It stores no personal data and returns no buyer or seller identities.
• Data is retrieved through a licensed data API, not by driving your account: no login, no scraping infrastructure, no anti-bot circumvention.
• Output is provided as-is for market research. Sold-listing samples are point-in-time observations of a public marketplace, not audited sales records, and not a guarantee of future prices or demand.
• You are responsible for using the output in line with the terms of the marketplaces you sell on and the law in your jurisdiction. eBay and Amazon are trademarks of their respective owners; this actor is not affiliated with or endorsed by either company.
• Questions or a bug? Open an issue on the actor's Issues tab; see the API tab for integration details.

SEO keywords

ebay product research, ebay sold prices, ebay sell through rate, ebay comps, sold price lookup, ebay sold listings, ebay market research, ebay price research tool, sell through rate calculator, ebay median sold price, dropshipping product research, ebay reseller tools, ebay demand analysis, ebay competitor analysis, ebay pricing data, ebay sales history, ebay completed listings, product sourcing research, ebay arbitrage tool, amazon to ebay spread, ebay sales velocity, ebay listing competition, ebay category research.

Built by doesaiknow.