Ebay Email Scraper

eBay Email Scraper

eBay Email Scraper is a public eBay research Actor. It searches public eBay result pages or inspects explicit item pages and returns seller email addresses only when the address is visibly published in the fetched public HTML. It does not log in, use private APIs, bypass access controls, or infer hidden contact data.

Actor page: eBay Email Scraper on Apify

Value and use cases

Use it for public seller-contact research, marketplace catalog research, and e-commerce lead qualification where the seller intentionally publishes a business email. The output records the public page and extraction method so downstream users can review provenance. Public email addresses can be personal data; use them only for a legitimate purpose and follow eBay terms, privacy law, and anti-spam rules.

What the Actor does not promise

The Actor does not discover every seller email, reveal private contact data, or guarantee that a listing contains an email. It only reports non-eBay email addresses found in bounded public HTML. A blocked page, an item without a public email, or an email excluded by customDomains produces a diagnostic or no normal row rather than a guessed address.

Input

The input is a JSON object. Unknown top-level fields are rejected. Omitted fields use the defaults below.

Field	Type and limits	Default	Behavior
startUrls	Array of at most 50 objects, each with an eBay HTTP(S) /itm/ url	[]	Explicit item pages inspected first, in input order. Duplicate URLs are removed.
keywords	Array of at most 10 non-empty strings, each at most 200 characters	[]	Search terms run after startUrls, in input order. Duplicate terms are removed.
market	us, uk, de, fr, it, es, ca, or au	us	Selects the eBay host for keyword searches.
location	Optional non-empty string, at most 160 characters	omitted	Passed to eBay search as the public location parameter.
customDomains	Array of at most 50 domains, such as outlook.com or @outlook.com	[]	Retains only emails whose normalized domain matches one of these domains.
maxResults	Safe integer from 1 to 500	100	Global cap on normal email rows across explicit URLs and searches.
maxPages	Safe integer from 1 to 10	3	Maximum result pages fetched per keyword.
maxListingsPerSearch	Safe integer from 1 to 50	10	Maximum listing cards inspected per keyword; it does not override the global maxResults.
timeoutMs	Safe integer from 5,000 to 120,000	30000	Per-request timeout.
maxBytes	Safe integer from 100,000 to 10,000,000	5000000	Maximum decompressed response size retained per request.
maxRetries	Safe integer from 0 to 3	1	Bounded retries for transient failures.
userAgent	Optional non-empty string, at most 300 characters	omitted	Transparent HTTP user-agent override.
proxyConfiguration	Standard Apify Proxy object; useApifyProxy is boolean and proxyUrls accepts at most 20 non-blank URLs	direct access	Enables Apify Proxy or custom proxy URLs.

Runnable input examples

Minimal keyword search:

{
  "keywords": ["juicer"]
}

Inspect one public item page:

{
  "startUrls": [
    {
      "url": "https://www.ebay.de/itm/365959456652"
    }
  ],
  "market": "de",
  "maxResults": 3,
  "maxPages": 1,
  "maxListingsPerSearch": 3
}

Combine explicit pages and keyword searches with a domain filter:

{
  "startUrls": [
    {
      "url": "https://www.ebay.com/itm/365959456652"
    }
  ],
  "keywords": ["vintage camera", "film camera"],
  "market": "us",
  "location": "10001",
  "customDomains": ["outlook.com"],
  "maxResults": 10,
  "maxPages": 1,
  "maxListingsPerSearch": 2
}

Bounded developer options with Apify Proxy:

{
  "keywords": ["public seller email"],
  "maxResults": 5,
  "maxPages": 1,
  "maxListingsPerSearch": 2,
  "timeoutMs": 20000,
  "maxBytes": 2000000,
  "maxRetries": 0,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyCountry": "US"
  }
}

Output

Normal dataset rows contain a public email observation. The exact fields are:

Field	Meaning
url	Public eBay item URL where the observation was made; required on every row.
scrapedAt	ISO timestamp for the observation; required on every row.
searchUrl	Public search URL that produced the listing, when the row came from a keyword search.
keyword, location, page	Search context, when applicable. page is the one-based eBay result page.
title, description	Public listing text captured from the result card or item page.
sellerName, sellerUrl	Visible seller identity and public seller URL, when present.
email, emailDomain	Lower-cased non-eBay email and its normalized domain.
context	search_result for a card-level email or item_page for an item-page extraction.
extractionMethod	public_search_card_text, public_html_mailto, public_html_visible_text, or public_html_mailto_and_visible_text.
marketplace	Resolved eBay marketplace code. An explicit item page is resolved from its final host; search rows use the selected market.
isPubliclyPublished, isGuessed	Provenance flags. Normal rows are always true and false.
error, errorCode	Present only on a diagnostic row; normal rows do not contain these fields.

Example normal search-card row:

{
  "url": "https://www.ebay.com/itm/123456789012",
  "searchUrl": "https://www.ebay.com/sch/i.html?_nkw=juicer&_pgn=1",
  "keyword": "juicer",
  "page": 1,
  "title": "Commercial juicer",
  "sellerName": "Example Seller",
  "email": "sales@example.org",
  "emailDomain": "example.org",
  "context": "search_result",
  "extractionMethod": "public_search_card_text",
  "marketplace": "us",
  "isPubliclyPublished": true,
  "isGuessed": false,
  "scrapedAt": "2026-09-08T10:00:00.000Z"
}

Example normal item-page row:

{
  "url": "https://www.ebay.de/itm/365959456652",
  "title": "Vintage watch",
  "sellerName": "Vintage Store",
  "sellerUrl": "https://www.ebay.de/str/vintagestore24",
  "email": "vintagestore24@outlook.com",
  "emailDomain": "outlook.com",
  "context": "item_page",
  "extractionMethod": "public_html_mailto_and_visible_text",
  "marketplace": "de",
  "isPubliclyPublished": true,
  "isGuessed": false,
  "scrapedAt": "2026-09-08T10:00:01.000Z"
}

Diagnostics

Diagnostics are explicit four-field dataset rows with url, error, errorCode, and scrapedAt. They do not contain guessed contact data. Common codes are MISSING_TARGETS, SOURCE_BLOCKED, and NO_PUBLIC_EMAILS; transport failures can also use the response code supplied by the public request layer, such as RESPONSE_TOO_LARGE.

Example empty-input diagnostic:

{
  "url": "https://www.ebay.com/",
  "error": "Provide at least one public eBay item URL or search keyword.",
  "errorCode": "MISSING_TARGETS",
  "scrapedAt": "2026-09-08T10:00:00.000Z"
}

Example blocked-source diagnostic:

{
  "url": "https://www.ebay.com/sch/i.html?_nkw=juicer&_pgn=1",
  "error": "The public source returned an access-control page.",
  "errorCode": "SOURCE_BLOCKED",
  "scrapedAt": "2026-09-08T10:00:02.000Z"
}

Pagination, filtering, deduplication, and source behavior

Request precedence and pagination

The Actor normalizes and deduplicates the input, inspects startUrls first, and then searches each keyword on the selected market host. Explicit URLs are not replaced by search results. Every explicit URL is eligible for one bounded public item-page inspection, while a keyword search requests pages 1 through maxPages and processes at most maxListingsPerSearch distinct cards for that keyword. Page order is source order; there is no user-configurable sort.

maxResults is a global output cap. It applies across explicit item pages, search-card emails, and item-page emails. A per-keyword listing bound and a page bound can reduce results before the global cap is reached. There is no startUrls-style pagination for explicit item pages and no input for arbitrary search URLs. Once the global cap is reached, later keywords are not requested.

Extraction, filtering, and deduplication

The parser reads public search-card and item-page HTML only. It removes script, style, noscript, SVG, and template content before visible-text extraction, accepts mailto: links and visible email-like text, excludes eBay/example/invalid/test domains, and keeps at most 20 candidate emails per page. customDomains is applied after normalization; when it is non-empty, only matching domains are retained. Search-card emails are emitted without a detail request when they pass the filter; cards with no retained card-level email, including cards whose visible email fails the domain filter, are inspected at their public item URL.

Normal rows are deduplicated by the normalized email plus listing URL. Explicit URLs are deduplicated after URL normalization, and the same email on two different listings remains two observations. No hidden seller contact, login-only page, private API response, or inferred address is used.

Proxy, blocked, empty, and partial runs

Direct public requests are the default. proxyConfiguration.useApifyProxy: true uses Apify Proxy; non-empty proxyUrls are also passed to Apify's proxy configuration. Proxy use is reported in OUTPUT_SUMMARY.proxyUsed but proxying does not bypass a login or access-control decision. The Actor warms a normal public cookie session and uses bounded retries; it does not use stealth or CAPTCHA evasion.

HTTP 403/429 responses and pages containing common challenge markers such as bm-verify or /_sec/verify become SOURCE_BLOCKED diagnostics. A listing without a retained public email becomes NO_PUBLIC_EMAILS. If normal rows were collected before a later search failure, those rows remain in the dataset and the summary status is COMPLETED_WITH_DIAGNOSTICS. With no normal rows, hasData is false. Responses exceeding maxBytes are not retained.

Run summary and downloads

The key-value store contains OUTPUT_SUMMARY with status, hasData, normalRecords, diagnosticRecords, inspectedListings, searchCardsProcessed, requestedItemUrls, requestedKeywords, marketplace, proxyUsed, source, and finishedAt. status is SUCCEEDED when no diagnostics were emitted and COMPLETED_WITH_DIAGNOSTICS otherwise. The legacy OUTPUT record remains available with datasetItems, diagnosticItems, and uniqueItems counts.

Example OUTPUT_SUMMARY:

{
  "status": "SUCCEEDED",
  "hasData": true,
  "normalRecords": 2,
  "diagnosticRecords": 0,
  "inspectedListings": 2,
  "searchCardsProcessed": 0,
  "requestedItemUrls": 2,
  "requestedKeywords": 0,
  "marketplace": "us",
  "proxyUsed": false,
  "source": "ebay-public",
  "finishedAt": "2026-09-08T10:00:03.000Z"
}

You can download the dataset in various formats such as JSON, HTML, CSV, or Excel.

Cost and limits

There is no fixed per-result fee. Apify compute and storage usage depends on response size, pages, item inspections, retries, and proxy choice. Keep maxPages, maxListingsPerSearch, maxResults, timeoutMs, and maxBytes bounded. Public-source availability and the presence of seller emails are not guaranteed.

Running the Actor

Apify Console

1. Open the Actor page.
2. Enter a valid JSON input in the Input tab or use the form fields.
3. Start the Actor and review normal and diagnostic rows in the Dataset tab.
4. Review OUTPUT_SUMMARY in the Key-value store and download the dataset in the format you need.

API and CLI

Start a run with the Apify API using the Actor ID ORn6S9ye40i2eLDpT and a JSON input body, or use the CLI from this directory:

npm install
apify call ORn6S9ye40i2eLDpT -p '{"keywords":["juicer"],"maxResults":5}'

Local development

npm install
npm run lint
npm test
apify validate-schema
apify run --purge --input-file test/inputs/default.json
npm run validate

The repository test inputs are bounded public probes, not proof that every eBay marketplace or seller publishes an email. A diagnostic-only result is truthful evidence of the current source response, not a normal-record guarantee. Do not commit generated storage/ output.

Troubleshooting

The dataset has only diagnostics

Check errorCode and OUTPUT_SUMMARY. NO_PUBLIC_EMAILS means no qualifying public address was found in the bounded pages. SOURCE_BLOCKED means eBay returned an access-control or challenge response. Try a smaller public item set, a different marketplace, or a compliant proxy configuration; do not add credentials or attempt to evade the source controls.

Search returns fewer rows than expected

The source may have fewer qualifying public emails than listings. Also check customDomains, maxListingsPerSearch, maxPages, and the global maxResults cap. Search-card results are deduplicated by email plus listing URL, and a listing without a qualifying card email receives one bounded item-page inspection.

A request is too large or too slow

Increase maxBytes or timeoutMs only within their documented bounds, reduce the page/listing limits, and keep maxRetries bounded. The Actor intentionally stops when a response exceeds maxBytes.

Privacy, legal, and affiliation

This Actor processes publicly displayed seller contact information. You are responsible for having a lawful basis, respecting eBay's terms and robots or access policies, complying with privacy and anti-spam laws, and honoring opt-outs. Do not use the Actor to collect private, login-only, or access-controlled contact data. This Actor is an independent community tool and is not affiliated with, endorsed by, or sponsored by eBay Inc.

For defects or contract questions, use the Issues tab on Apify and include the input shape, run ID, summary, and diagnostic code without sharing private data.