Flippa Scraper — Online Businesses & Websites For Sale

Every listing on Flippa, the largest marketplace for buying and selling online businesses: e-commerce stores, content sites, SaaS, Amazon FBA storefronts, iOS and Android apps, and domains.

Asking price, monthly revenue and profit, traffic, verification flags, auction state and seller location. No API key, no login.

---

What you get

Three record types share one dataset, told apart by recordType:

• LISTING — title and summary, the Flippa URL, propertyType, price, status and saleMethod, the financials Flippa surfaces (ttm_revenue, revenue_multiple, profit_average, multiple, monetization), auction state (bid_count, end_at), when it was established, the country and primary platform, the verification flags, and a thumbnail. Flippa's own field names are passed through verbatim, so any field it adds appears without an actor update.
• SEARCH_SUMMARY — one per run: filters applied, upstream's total and whether it is exact, pages and requests made, rate-limit hits.
• ERROR — one per failed run, so every run produces at least one row.

Every row carries _input, _source, _scrapedAt, recordType.

Financials are seller-reported — and the flags say when they aren't

revenuePerMonth and profitPerMonth are whatever the seller entered. Flippa separately verifies some listings, so hasVerifiedRevenue and hasVerifiedTraffic travel beside those numbers on every row and are never null — "unverified" is a real state, not missing data. On a sample of 100 listings, 5 carried verified revenue. Filter on those flags before treating any of it as fact.

---

Two things worth knowing before you read the numbers

The total clamps at 10,000. Below that it is a real count and paging ends exactly on it — website reports 4,627, fba 281, ios_app 247. At 10,000 it is a floor, not a count: a very broad category such as domain reports exactly 10,000. totalIsExact tells you which case you are in. Narrow by type or price to get a real count.

propertyTypes are categories, not row types. website covers several display types, so asking for it returns rows whose propertyType reads Ecommerce, Content, SaaS, Marketplace, Service or Marketing Agency. That is correct, not a broken filter. The summary row carries both displayTypesExpected and displayTypesSeen so you can check the filter bound without knowing Flippa's taxonomy.

An empty result stays empty. If a query matches nothing, Flippa's page shows a "you might like" promo strip; the actor recognises that (the reported total is 0) and returns no LISTING rows for it, rather than passing the promo listings off as matches.

---

Speed and the proxy

The /search page the actor reads is only lightly rate-limited — a burst of requests from one IP comes back fine — so runs are fast even without a proxy.

The Apify residential proxy (on by default) takes a fresh exit IP per request, which removes any limit you might hit on a very large run; turning it off simply paces a little more conservatively. rateLimitHits on the summary row tells you whether a limit was ever met (it is normally 0).

---

Input

Field	Meaning
propertyTypes	Flippa's six categories: website, fba, ios_app, android_app, domain, other. Empty for everything.
status	any, or open for currently live listings — approximate, see below.
saleMethod	any, auction, or classified (fixed price).
minPrice / maxPrice	Bounds on the listing price, in USD.
maxItems	Default 200; 0 for everything the query can reach.
pageSize	100 max and the sensible default — Flippa limits by request, so fewer, larger pages is strictly better.
minRequestInterval	Leave unset; the actor picks a light default (0.5s with a proxy, 1.5s without).
proxyConfiguration	Residential, on by default. See above.

---

Known limits

• 10,000 is the highest total Flippa will report. The listings beyond it exist but the count does not distinguish them.
• Seller-reported financials unless the verification flags say otherwise.
• Only the fields Flippa exposes on its listings API — no traffic analytics, and no seller contact details.
• status=open is approximate. Flippa's status filter lags the live listing state: measured twice over the same 290-result set, ~5 rows came back completed, clustered on later pages. Nothing is dropped or rewritten — every row carries its own authoritative status, and the summary reports statusFilterIsApproximate and statusesSeen. Filter on the row if you need exactness.
• A passed leaf type (e.g. ecommerce_store) is refused with an explanation; pass the category (website) instead.
• bizbuysell.com and bizquest.com are deliberately not covered. Both disallow CCBot, which this portfolio treats as part of the disqualifying AI-bot set.