eBay Seller Feedback Records

Export ebay seller feedback as structured records for seller due diligence and recurring reputation analysis. Supply public seller usernames or ebay.com seller profile URLs and receive feedback IDs, ratings, comments, coarse time labels, masked reviewer aliases and public item transaction context.

This Actor collects feedback received as seller, not product reviews or feedback left as a buyer. It includes eBay automated feedback and labels it so downstream analysis can distinguish automatic entries from buyer-written comments. No login or private account is required.

Who is it for?

Marketplace analysts can compare recent feedback across candidate sellers. Procurement teams can inspect complaints before choosing a supplier. Seller operations teams can schedule snapshots and detect newly appearing feedback IDs. Researchers can join public item context to sentiment analysis without attempting to identify masked reviewers.

Why use it?

Stable source feedback IDs support deduplication across snapshots. Seller-only filtering avoids mixing seller reputation with purchases made by that seller. Public item prices remain strings with source currency rather than misleading normalized numbers. Coarse time labels remain coarse: the Actor does not invent exact dates.

Get started

1. Add seller usernames or public ebay.com seller-profile URLs.
2. Set the global feedback limit, from 1 to 1000.
3. Start the Actor and wait for completion.
4. Inspect the default dataset, then export JSON, CSV or Excel through Apify.
5. For repeat use, save your input as a Task and add an Apify schedule yourself.

{"sellers":["musicmagpie"],"maxItems":30}

Inputs

Input	Meaning
sellers	Required array of 1–20 usernames or HTTPS ebay.com /usr/USERNAME and /fdbk/feedback_profile/USERNAME URLs. Duplicate sellers are merged case-insensitively.
maxItems	Global maximum unique feedback records across all sellers in input order. Default 100; allowed 1–1000. Zero is rejected, not unlimited.

Only ebay.com is supported. Links to item pages, country-specific domains and arbitrary websites are rejected. URL query parameters do not override the Actor's seller-only scope. Collection uses source time order and includes automated feedback. There are no date, text, polarity or locale filters.

Extracted fields

Field	Meaning
sellerUsername, sellerProfileUrl	Normalized seller and public feedback profile
feedbackId	Stable source feedback ID; combine with seller for joins
rating	Positive, neutral or negative
comment	Public comment, possibly empty for automated entries
timeLabel	Source label such as Past month; never an exact date
reviewerAlias	Public masked alias only; null if not masked
automatedFeedback	Source identifies eBay automated feedback
itemId, itemUrl, itemTitle	Public transaction item context where supplied
transactionPrice	Displayed amount and currency, without conversion
verifiedPurchase	Source displays the verified-purchase signal
sourceUrl, scrapedAt	Endpoint/page provenance and UTC extraction timestamp

Unavailable optional context is null. A missing alias does not imply a missing reviewer. The Actor never attempts to resolve identities hidden by eBay.

Output example

The following preserves the shape of an observed record while anonymizing identifiers and content:

{
  "sellerUsername":"example-seller",
  "sellerProfileUrl":"https://www.ebay.com/fdbk/feedback_profile/example-seller",
  "feedbackId":"3051000000001",
  "rating":"positive",
  "comment":"Item arrived as described.",
  "timeLabel":"Past month",
  "reviewerAlias":"a***b",
  "automatedFeedback":false,
  "itemId":"301700000001",
  "itemUrl":"https://www.ebay.com/itm/301700000001",
  "itemTitle":"Example paperback book",
  "transactionPrice":"GBP 3.48",
  "verifiedPurchase":true,
  "sourceUrl":"https://www.ebay.com/fdbk/update_feedback_profile",
  "scrapedAt":"2026-10-24T14:00:00.000Z"
}

How much does it cost to export eBay seller feedback?

Pay per event: one start charge of $0.001 per run, plus one item charge for each unique delivered feedback record. No separate charge applies to item context, automated-feedback flags or run metrics. Invalid inputs are rejected before the start event. An upstream failure after starting can still incur the start charge and charges for any already delivered records.

Store spend tier	Price per feedback record
FREE	$0.00092
BRONZE	$0.0008
SILVER	$0.000624
GOLD	$0.00048
PLATINUM	$0.00048
DIAMOND	$0.00048

At BRONZE, 10 records cost an estimated $0.009, 100 cost $0.081, and 1000 cost $0.801 including one start event. Spend tiers reflect your qualifying aggregate monthly Apify Store spend, not volume in this Actor alone. Applicable Apify account charges and tax may differ. No customer-supplied proxy or paid API account is required.

Pagination and limitations

Requests are sequential with brief spacing, in pages of 100. Collection stops at the global cap or source exhaustion. Rows are deduplicated by seller plus feedback ID. A repeated full page fails rather than pretending pagination finished. Source updates during collection can still shift the window: this is a snapshot, not a transactional or complete historical archive.

The source may block requests, change its response or hide older records. Upstream errors fail honestly; rows already written remain partial results. A failed run must not be treated as a complete snapshot. Do not infer the absence of complaints from a capped result. This Actor does not provide alerts, historical storage, product-review collection, identity enrichment or continuous monitoring by itself.

Integrations

Send dataset exports to Google Sheets for a seller comparison, load JSON into a warehouse, or use an Apify webhook to trigger your processing after success. For recurring monitoring, schedule a saved Task and upsert on (sellerUsername, feedbackId); preserve scrapedAt as observation time. Compute your own new-record delta and complaint summaries. Automated entries should be excluded from human-sentiment analysis unless your methodology intentionally includes them.

API usage

Keep your Apify token in an environment variable, never in source or public URLs.

curl -X POST 'https://api.apify.com/v2/acts/automation-lab~ebay-seller-feedback-records/runs' \
  -H "Authorization: Bearer $APIFY_TOKEN" -H 'Content-Type: application/json' \
  -d '{"sellers":["musicmagpie"],"maxItems":30}'

The asynchronous response contains run and storage IDs, not the feedback rows. Follow that run until terminal status, then read its default dataset.

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('automation-lab/ebay-seller-feedback-records')
  .call({ sellers: ['musicmagpie'], maxItems: 30 });
if (run.status !== 'SUCCEEDED') throw new Error(`Run ${run.id}: ${run.status}`);
const { items } = await client.dataset(run.defaultDatasetId).listItems({ limit: 30 });
console.log(items);

import os
from apify_client import ApifyClient
client = ApifyClient(os.environ['APIFY_TOKEN'])
run = client.actor('automation-lab/ebay-seller-feedback-records').call(
    run_input={'sellers': ['musicmagpie'], 'maxItems': 30})
if run['status'] != 'SUCCEEDED':
    raise RuntimeError(run['status'])
rows = client.dataset(run['defaultDatasetId']).list_items(limit=30).items

MCP use

Select this stable Actor subset:

claude mcp add --transport http apify \
  "https://mcp.apify.com?tools=automation-lab/ebay-seller-feedback-records"

Claude Desktop, Cursor, and VS Code configuration (use your client's supported remote HTTP MCP configuration):

{"mcpServers":{"apify":{"url":"https://mcp.apify.com?tools=automation-lab/ebay-seller-feedback-records"}}}

Authenticate through your client's supported Apify authentication. Discover actual tool names and input shapes with scoped tools/list. Actor selection also adds run/data tools; it does not expose exactly one tool. Discovery is read-only and does not authorize execution.

Example prompt for an authorized MCP request: “Collect 30 feedback records for musicmagpie, then summarize negative feedback without identifying masked reviewers.” Start once, retain the run ID and storage IDs. call-actor waits at most 45 seconds (waitSecs, default 30), so a nonterminal response is not failure or output. Follow the same run using get-actor-run with 2/4/8-second backoff capped at 10 seconds and a 120-second total deadline. Set the client timeout above the server wait plus transport margin. On timeout recover the known run; never launch a duplicate merely to poll. If the run ID is unknown, report uncertainty. At deadline report pending and stop.

After success, read get-dataset-items using the returned dataset ID and explicit fields, offset and limit according to the discovered schema. An example consumer budget is pages of 20 source rows, at most 100 rows and 64 KiB of serialized UTF-8 content admitted to model context, whichever is reached first. Enforce the byte ceiling host-side, including oversized rows, and disclose omissions or summaries. Without client-side interception no hard byte guarantee is possible. Track source offsets rather than transformed counts. Report rows returned, continuation offset and partial/truncated status. Keep full exports in storage outside model context. These are consumer-context limits, not Actor collection limits. No universal token-saving claim is made.

Legality and responsible use

Use only public data for legitimate analysis consistent with applicable law, privacy obligations and eBay terms. Public availability does not grant unrestricted reuse. This independent Actor is not affiliated with or endorsed by eBay or Apify. It uses Apify's standard user terms and no additional custom terms.

Data handling and retention

The Actor does not use AI, private credentials or an external paid data supplier. It processes public comments, masked aliases and item context, which may still contain personal information written by users. Avoid attempts to identify people. Only masked aliases are exported; unmasked reviewer names are omitted. Feedback content is not sanitized into a guarantee of anonymity.

Data and logs are stored in your Apify run storage under your account's retention settings. Delete runs and associated datasets/key-value stores when no longer needed; the Actor creates no independent cross-run database or cache. Run metrics contain counts only. Failed operations send sanitized diagnostic input, exceptions and actor/build/run IDs to our private GlitchTip service for repair; secret fields and URL queries are removed and reports are retained for 30 days. No raw response body is sent deliberately. Apify hosts execution/storage; our private GlitchTip receives failure diagnostics. No model provider receives scraped data.

Troubleshooting and FAQ

Why did the run fail after saving rows? An upstream page can become unavailable. Saved rows are partial. Use the run status, not just a nonempty dataset, to assess completion. Repeated 429/temporary server failures receive bounded backoff; challenges are not bypassed or reported as an empty success.

Why is a date only “Past month”? That is the public source's precision. scrapedAt is observation time, not transaction time.

Can I filter out automatic feedback? Not at collection time. Use automatedFeedback in downstream processing; automatic rows count toward the record limit and price.

Can I scrape item reviews or UK/German profiles? No. The supported contract is feedback received as seller via ebay.com, including public transaction context when available.

Can I collect all history? No. The 1000-row global cap and upstream availability apply. Pagination does not promise an exhaustive archive.

Where can I get support? Open an issue on this Actor's Apify Store page with your run link, expected result and a non-sensitive input example. Do not post tokens, private data or credentials. User reports receive a substantive response within 14 calendar days.

Related Actors

Use eBay Scraper for listing research rather than seller-feedback records. Join only on public item IDs when present; missing item context cannot be reconstructed by this Actor.