eBay Product Details Scraper With Buyer Reviews & Ratings

eBay Scraper — Product Details, Listing & Seller-Wide Feedback

This actor turns eBay listing URLs into full product records plus the buyer feedback behind them — rating, comment text, verified-purchase flag, buyer photos and time frame — each entry labelled as feedback about that exact listing or about the seller overall. Every response is structured JSON, ready to pass directly to an LLM, index into a vector store, or feed a monitoring pipeline. Paste one URL or a thousand and get one product row plus one row per feedback entry back in the same run.

What is eBay Product Details Scraper With Buyer Reviews & Ratings?

eBay Product Details Scraper With Buyer Reviews & Ratings reads eBay listing pages by URL and returns two things in one dataset: the full product record (title, price, condition, images, specifics, description, shipping and return policy) and the buyer feedback behind it, labelled by scope — this_item or seller_all_items. No eBay account or login is used anywhere in the run; every page it reads is the same public listing a browser would load. What sets it apart from a plain listing scraper is that scope label: feedback about the exact item you asked about is never mixed in with the seller's much larger history across everything else they sell.

Key capabilities:

• 🛍️ Full listing detail — title, price, condition, images, item specifics, description, shipping, delivery and return policy
• 💬 Buyer feedback rows — one per feedback entry, interleaved right after the listing it belongs to
• 🎯 Scope labelling — every feedback row states whether it is about this listing or the seller overall
• 🔎 Filtering — by rating, verified-purchase status, or presence of buyer photos
• 🤖 Optional AI tagging — sentiment, complaint category, and a description-versus-condition mismatch check, off by default
• 🌍 Any eBay marketplace — .com, .co.uk, .de, .com.au and the rest, inferred from the URL itself

What data can you get with eBay Product Details Scraper With Buyer Reviews & Ratings?

Every run returns product records and, when feedback collection is on, one entry per buyer feedback item — plus optional AI-derived labels on top of both.

Result Type	Extracted Fields	Primary Use Case
Product / listing record	product_title, product_price, product_condition, product_image, item_specifics, item_description, shipping_info, delivery_info, return_policy, item_number, detail_url, seller_name, seller_username, seller_items_sold, positive_feedback	Full listing detail for sourcing, repricing or a competitor teardown
Buyer feedback entry	reviewScope, reviewerName, reviewerFeedbackScore, ratingType, reviewComment, isVerifiedPurchase, isAutomatedFeedback, reviewDateRange, reviewPhotoUrls, reviewPhotoCount	Reading what buyers actually said, filtered by rating or verified-purchase status
AI enrichment (optional)	aiSentiment, aiComplaintType, aiConditionMismatch, aiConditionMismatchNote	Bulk-labelling feedback text and catching description-vs-condition contradictions without reading every comment by hand

Listing-scoped vs. seller-wide feedback

Every feedback row is tagged with exactly where it came from: this_item for buyers who bought this exact listing, seller_all_items for feedback left about the same seller across all of their other listings. Many eBay data tools return one combined feedback count and stop there; this actor keeps the two apart on every row through reviewScope, so this_item text is never diluted with feedback about a completely different item the same seller happens to sell. With feedbackSource: "both" (the default), listing-specific entries are always exhausted first and seller-wide entries only fill whatever budget in maxReviewsPerItem is left over — so the most relevant text for the listing you asked about is never crowded out by the seller's much larger, less specific feedback history.

{
  "reviewScope": "this_item",
  "reviewerName": "k***c",
  "reviewerFeedbackScore": 970,
  "ratingType": "Positive",
  "reviewComment": "Arrived brand new, not refurbished as described — great surprise.",
  "isVerifiedPurchase": true,
  "reviewDateRange": "Past month"
}

AI condition-mismatch detection

When aiReviewTagging is on and a model key is supplied, each listing's free-text description is checked against its declared product_condition — for example, a listing marked New whose description mentions scratches or missing parts. A contradiction sets aiConditionMismatch: true with a one-sentence aiConditionMismatchNote; no contradiction sets it false. Each feedback comment is separately labelled with aiSentiment (positive/neutral/negative/mixed) and aiComplaintType (one of a fixed set including item_not_as_described, damaged_or_defective, shipping_delay and others). Without a key, all of these columns stay null rather than guessed.

How does eBay Product Details Scraper differ from eBay's official APIs?

eBay's own Developer Program exposes the Browse API for item lookups and a separate Feedback API for managing an account's own feedback; eBay Product Details Scraper With Buyer Reviews & Ratings covers similar ground for public listings, without the registration step.

Feature	eBay Developer APIs (Browse + Feedback, developer.ebay.com, checked 2026-07-30)	eBay Product Details Scraper With Buyer Reviews & Ratings
Setup	App registration, OAuth client credentials, eBay Developer Program approval	Paste listing URLs into the input and run — no eBay account, no app registration
Item detail retrieval	Browse API getItem / getItemByLegacyId return structured item data	Same core listing detail (title, price, condition, images, specifics, description) read from the public page
Feedback for a specific listing, any seller	The Feedback API's getFeedback returns line items and feedback tied to the authenticated account's own transactions, by that account's role — not another seller's feedback	Reads any public listing's feedback for any seller, with no need to authenticate as that seller
Feedback scope labelling	Not documented as a resource of the public Feedback API	Every feedback row carries reviewScope: this_item or seller_all_items
Output shape	One schema per endpoint, called separately for item data and for feedback	One flattened JSON row per product or feedback entry, in a single dataset
Access limits	Subject to eBay Developer Program call quotas	Subject to the run's own proxy and duration settings, not a developer-program quota

The official APIs are the right choice when you already operate the eBay seller or buyer account whose own feedback and orders you need, inside an authenticated integration. This actor is the more direct route when you want product and feedback data for listings you don't hold an account for, without registering an app first.

What is the difference between eBay "reviews" and buyer feedback?

eBay does not run a 5-star product review system on ordinary seller listings — what shoppers search for as "eBay reviews" is actually eBay feedback, a Positive / Neutral / Negative rating plus an optional comment a buyer leaves after a transaction. The distinction matters because the two behave completely differently: a star rating (as used on Amazon) aggregates opinions about a product regardless of who sold it, while eBay feedback is tied to a specific seller and, on the listing page, split further into feedback about this exact item versus the seller's history across everything they sell. Confusing the two leads to bad expectations — there is no 1-to-5 score to scrape, and a seller's overall percentage can look excellent while a specific listing's this_item feedback tells a different story. eBay Product Details Scraper With Buyer Reviews & Ratings returns both scopes explicitly: ratingType carries eBay's real Positive/Neutral/Negative label, never converted into a fake star score, and reviewScope tells you whether each entry belongs to this_item or seller_all_items.

How to scrape eBay listings with this actor

1. Open eBay Product Details Scraper With Buyer Reviews & Ratings on its Apify Store listing and click Try for free (or Start, if you already use it).
2. Paste one or more eBay listing addresses into reviewSourceUrls — copied straight from the browser address bar, e.g. https://www.ebay.com/itm/157967038228.
3. Set the feedback controls that matter for the job: feedbackSource (both / thisItemOnly / sellerWide), ratingFilter (all / positive / neutral / negative), verifiedPurchaseOnly, withBuyerPhotosOnly and maxReviewsPerItem.
4. Click Start and watch rows appear in the Output tab as each listing finishes — nothing waits for the whole run to complete.
5. Export the dataset as JSON, CSV, Excel or XML from the Output tab, or pull it through the Apify API into your own pipeline.

{
  "reviewSourceUrls": ["https://www.ebay.com/itm/157967038228"],
  "feedbackSource": "both",
  "maxReviewsPerItem": 25,
  "ratingFilter": "all"
}

How to scrape multiple eBay listings in one job

Add every listing address to the same reviewSourceUrls array — one URL per line in the editor — and they all run inside a single job. The actor works through them one at a time, in the order supplied, with a short randomized pause between listings; there is no separate batch endpoint or upload format, the array itself is the batch.

⬇️ Input

Every field is optional; supply at least one listing address in reviewSourceUrls (or one of its legacy aliases below).

Parameter	Required	Type	Description	Example Value
reviewSourceUrls	No	array of strings	One eBay listing address per line, copied from the browser address bar (e.g. https://www.ebay.com/itm/123456789012). Works with any eBay marketplace — .com, .co.uk, .de, .com.au and the rest. Add as many as you like; each becomes its own product row.	["https://www.ebay.com/itm/157967038228"]
urls	No	array of strings	Legacy alias kept so an input written for the original eBay Product Details Scraper keeps running unchanged. Used only when reviewSourceUrls is empty.	[]
startUrls	No	array of strings	Second legacy alias for the same list. Used only when both reviewSourceUrls and urls are empty.	[]
includeBuyerFeedback	No	boolean	Default true. On: every listing row is followed by one row per buyer feedback entry, so comment text sits next to the product it belongs to. Off: only product rows (plus feedback counters) are returned.	true
feedbackSource	No	string (enum)	Default "both". "both" — this listing first, then seller-wide entries fill the remaining budget. "thisItemOnly" — strictly entries left after buying this exact listing. "sellerWide" — entries left for the seller across all of their listings.	"both"
maxReviewsPerItem	No	integer	Default 20, minimum 0, maximum 500. Caps the EXTRA feedback rows per listing; product rows are never limited by this number. 0 = as many as eBay will serve, up to the hard ceiling of 500.	25
ratingFilter	No	string (enum)	Default "all". "all", "positive", "neutral", "negative". Negative-only is the fastest way to surface complaints; positive-only is useful for testimonial text.	"all"
verifiedPurchaseOnly	No	boolean	Default false. Drops any feedback entry eBay does not mark as a verified purchase.	false
withBuyerPhotosOnly	No	boolean	Default false. Keeps only feedback entries where the buyer attached at least one photo, and returns the photo addresses on the row.	false
aiReviewTagging	No	boolean	Default false. When on, each feedback comment gets a sentiment label and complaint category, and each listing gets a description-versus-condition contradiction check. Needs a model key below; without one the AI columns stay empty and everything else still runs.	false
aiModel	No	string (enum)	Default "claude-haiku-4-5". One of: claude-haiku-4-5, claude-sonnet-5, claude-opus-4-8, gpt-4o-mini, gpt-4o, gpt-4.1-mini, gpt-4.1, o3-mini, gemini-2.0-flash-lite, gemini-2.0-flash, gemini-2.5-flash, gemini-2.5-pro, grok-3-mini, grok-3, deepseek-chat, sonar, mistral-small-latest, mistral-large-latest. Provider is inferred from the model name.	"claude-haiku-4-5"
aiApiKey	No	string (secret)	Your own key for the provider chosen above. If left empty, falls back to an environment variable: ANTHROPIC_API_KEY, OPENAI_API_KEY, GOOGLE_API_KEY, XAI_API_KEY, DEEPSEEK_API_KEY, PERPLEXITY_API_KEY or MISTRAL_API_KEY. With no key at all, AI columns are returned empty rather than guessed.	""
proxyConfiguration	No	object	Default { "useApifyProxy": true }. The actor starts on the fastest available connection and escalates to a steadier one automatically if a listing does not come through.	{ "useApifyProxy": true }

Input example

{
  "reviewSourceUrls": [
    "https://www.ebay.com/itm/157967038228",
    "https://www.ebay.co.uk/itm/210987654321"
  ],
  "includeBuyerFeedback": true,
  "feedbackSource": "both",
  "maxReviewsPerItem": 25,
  "ratingFilter": "all",
  "verifiedPurchaseOnly": false,
  "withBuyerPhotosOnly": false,
  "aiReviewTagging": false,
  "proxyConfiguration": { "useApifyProxy": true }
}

Common pitfall: maxReviewsPerItem: 0 does not mean unlimited — it means "as many as eBay will serve," which this actor caps at a hard ceiling of 500 feedback entries per listing. Raising it above 500 has no effect either; the schema itself refuses values past that maximum.

⬆️ Output

Every run writes to one Apify dataset. Two row shapes share it, told apart by type and isChild, so either one can be filtered out with a single dataset query. Export it as JSON, CSV, Excel or XML from the Output tab, or read it through the Apify API.

Charged events: row_result, fired once per product row and once per buyer feedback row. Rows for a listing that could not be read at all — not an eBay URL, blocked, or removed — are pushed without that charge. Filter on success: true to see only billed rows, or keep success: false rows if you'd rather see which listings failed instead of having them silently disappear.

Product row (type: "product", isChild: false) carries, on success: seller_name, seller_username, itemFeedbackCount, sellerFeedbackCount, seller_items_sold, positive_feedback, product_title, product_image, product_price, product_condition, shipping_info, delivery_info, return_policy, item_number, item_description, item_specifics, detail_url, type, isChild, source_url, success, proxyMode, buyerFeedbackCollected, itemScopedFeedbackCollected, sellerScopedFeedbackCollected, duplicateFeedbackCollapsed, aiConditionMismatch, aiConditionMismatchNote, scrapedAt. proxyMode records which connection tier served the page — direct, datacenter or residential.

When a listing cannot be read at all, a shorter row is pushed instead: type, isChild, source_url, success: false, error, scrapedAt, and proxyMode (failed or not_found) when the failure happened after a fetch attempt.

Buyer feedback row (type: "buyer_feedback", isChild: true) carries: type, isChild, parentItemNumber, parentProductTitle, parentSourceUrl, seller_username, reviewIndex, reviewScope, reviewerName, reviewerFeedbackScore, ratingType, reviewComment, isVerifiedPurchase, isAutomatedFeedback, reviewDateRange, reviewPhotoUrls, reviewPhotoCount, aiSentiment, aiComplaintType, success, scrapedAt.

Scraped results

[
  {
    "type": "product",
    "isChild": false,
    "seller_name": "Verdi Commerce",
    "seller_username": "verdi_commerce",
    "itemFeedbackCount": 22,
    "sellerFeedbackCount": 64521,
    "seller_items_sold": "12.4K",
    "positive_feedback": "98.5% positive",
    "product_title": "Netgear Orbi 970 Quad-Band WiFi 7 Mesh System",
    "product_image": ["https://i.ebayimg.com/images/g/abcAAOSwexample/s-l1600.jpg"],
    "product_price": "$1,299.99",
    "product_condition": "Excellent - Refurbished",
    "shipping_info": "Located in: Austin, TX, United States",
    "delivery_info": "Fri, Aug 1 and Wed, Aug 6",
    "return_policy": "30 days returns. Buyer pays for return shipping.",
    "item_number": "136689000538",
    "item_description": "Fully tested, includes original power adapters and satellite units.",
    "item_specifics": { "Brand": "NETGEAR", "Model": "Orbi 970", "Connectivity": "Wi-Fi 7" },
    "detail_url": "https://www.ebay.com/itm/136689000538",
    "source_url": "https://www.ebay.com/itm/136689000538",
    "success": true,
    "proxyMode": "datacenter",
    "buyerFeedbackCollected": 2,
    "itemScopedFeedbackCollected": 2,
    "sellerScopedFeedbackCollected": 0,
    "duplicateFeedbackCollapsed": 3,
    "aiConditionMismatch": null,
    "aiConditionMismatchNote": null,
    "scrapedAt": "2026-07-30T12:00:00Z"
  },
  {
    "type": "buyer_feedback",
    "isChild": true,
    "parentItemNumber": "136689000538",
    "parentProductTitle": "Netgear Orbi 970 Quad-Band WiFi 7 Mesh System",
    "parentSourceUrl": "https://www.ebay.com/itm/136689000538",
    "seller_username": "verdi_commerce",
    "reviewIndex": 1,
    "reviewScope": "this_item",
    "reviewerName": "k***c",
    "reviewerFeedbackScore": 970,
    "ratingType": "Positive",
    "reviewComment": "They said it was used and refurbished. It arrived BRAND NEW! Great savings.",
    "isVerifiedPurchase": true,
    "isAutomatedFeedback": false,
    "reviewDateRange": "Past month",
    "reviewPhotoUrls": [],
    "reviewPhotoCount": 0,
    "aiSentiment": null,
    "aiComplaintType": null,
    "success": true,
    "scrapedAt": "2026-07-30T12:00:00Z"
  },
  {
    "type": "buyer_feedback",
    "isChild": true,
    "parentItemNumber": "136689000538",
    "parentProductTitle": "Netgear Orbi 970 Quad-Band WiFi 7 Mesh System",
    "parentSourceUrl": "https://www.ebay.com/itm/136689000538",
    "seller_username": "verdi_commerce",
    "reviewIndex": 2,
    "reviewScope": "this_item",
    "reviewerName": "m***a",
    "reviewerFeedbackScore": 214,
    "ratingType": "Negative",
    "reviewComment": "One antenna arrived cracked and the seller was slow to respond about a return.",
    "isVerifiedPurchase": true,
    "isAutomatedFeedback": false,
    "reviewDateRange": "Past 6 months",
    "reviewPhotoUrls": ["https://i.ebayimg.com/images/g/fdbkexample/s-l400.jpg"],
    "reviewPhotoCount": 1,
    "aiSentiment": null,
    "aiComplaintType": null,
    "success": true,
    "scrapedAt": "2026-07-30T12:00:00Z"
  }
]

⚠️ A few honest limits before relying on this data: eBay caps how much feedback exists to read, and this actor never exceeds the hard ceiling of 500 entries per listing even when maxReviewsPerItem is 0. eBay also masks buyer names itself — k***c is exactly what the page shows, nothing is redacted further here — and it publishes a time frame rather than an exact date (reviewDateRange carries "Past month" or similar, verbatim). Repeated automated delivery-confirmation text is collapsed into a single row rather than counted twice, and reported honestly via duplicateFeedbackCollapsed — which is why buyerFeedbackCollected can be lower than itemFeedbackCount on a listing.

How can I use the data from eBay Product Details Scraper With Buyer Reviews & Ratings?

• 🛍️ Sourcing and repricing teams: pull product_price, product_condition and item_specifics for a list of item numbers to compare condition-adjusted pricing before buying or relisting stock.
• 🤖 AI engineers and LLM developers: an agent issues a listing URL, receives the structured JSON row back, and passes product_title, item_specifics and reviewComment values straight into the model as grounding context — no scraping code inside the agent itself.
• 📊 Market and competitive researchers: benchmark a competitor's listing by reading what this_item feedback actually praises or complains about, separately from the seller's much larger seller_all_items history.
• 🛠️ Quality and support teams: filter ratingFilter: "negative" and verifiedPurchaseOnly: true to export only confirmed complaints, and use aiComplaintType to group them by category without reading every comment by hand.

How do you monitor eBay seller feedback and listing changes over time?

Tracking a seller's feedback quality, or a listing's price and condition, is a repeated-query problem: one run tells you where things stand today; only a second run against the same reviewSourceUrls tells you what changed. Because this actor takes the same set of listing URLs on every run, scheduling it on a cadence and diffing the output against the previous run's dataset turns a one-off lookup into ongoing monitoring.

Fields worth diffing between runs: product_price and product_condition (has the seller changed the listing), itemFeedbackCount and sellerFeedbackCount (raw counters straight off the page), buyerFeedbackCollected and the set of reviewComment values (new feedback text that wasn't there last time), and — with AI tagging on — aiSentiment and aiComplaintType (has the mix of complaints shifted). duplicateFeedbackCollapsed shows how much of eBay's own repeated automated text was filtered out before you even get to the diff.

A practical loop: save the listing URLs you care about once, run the actor on an Apify Schedule (daily or weekly), keep each run's dataset, and compare the new reviewComment values against the previous run's set — alert when a new negative ratingType entry appears, or when aiConditionMismatch flips to true on a listing that previously matched its description.

Integrate eBay Product Details Scraper With Buyer Reviews & Ratings and automate your workflow

eBay Product Details Scraper With Buyer Reviews & Ratings works with any language or tool that can send an HTTP request — it runs on the Apify platform, so the same run can be started from the Console, the Apify API, or any of Apify's official client libraries.

REST API with Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")

run = client.actor("<your-username>/ebay-product-details-scraper-with-buyer-reviews-and-ratings").call(
    run_input={
        "reviewSourceUrls": ["https://www.ebay.com/itm/157967038228"],
        "feedbackSource": "both",
        "maxReviewsPerItem": 25,
    }
)

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    label = item.get("product_title") or item.get("reviewComment")
    print(item["type"], label)

Scheduled monitoring and delivery

Apify Schedules re-run this actor on a cron-style interval against the same reviewSourceUrls without any code of your own. Point an Apify Webhook at your endpoint for the run-succeeded event to receive each run's dataset the moment it finishes, instead of polling for it.

⚖️ Is it legal to scrape eBay listings and feedback?

Yes — reading publicly visible eBay listing and feedback pages is legal; eBay Product Details Scraper With Buyer Reviews & Ratings collects only what any visitor sees in a browser, without logging in or bypassing any access control. Product data — titles, prices, condition, specifics — is business/catalogue data governed primarily by eBay's own Terms of Service and, in some jurisdictions, database-rights law, rather than personal-data law. Buyer feedback rows carry a reviewer's eBay username (already masked by eBay itself, e.g. k***c) and comment text; because a username can act as an identifier tied to a person, treat feedback data under GDPR/CCPA principles if you store or reuse it — minimize what you keep and avoid trying to re-identify buyers. A one-off competitive check carries a different risk profile than storing feedback in bulk for AI training. Consult your legal team for commercial use cases involving bulk data storage.

❓ Frequently asked questions

Does this eBay scraper work on ebay.co.uk, ebay.de or other eBay marketplaces?

Yes — paste a link from any eBay marketplace domain (.com, .co.uk, .de, .com.au and the rest) into reviewSourceUrls; the actor detects the marketplace from the URL itself and reads feedback from that same domain, not a hardcoded one.

Can I collect feedback for just this listing, or just the seller's overall history?

Yes. Set feedbackSource to thisItemOnly for strictly the entries left after buying this exact listing, or sellerWide for the seller's feedback across everything they sell. The default, both, exhausts listing-specific entries first and only uses seller-wide entries to fill whatever budget in maxReviewsPerItem is left.

How does the actor handle eBay's anti-bot measures?

It fetches pages with a Chrome-impersonating HTTP client and starts on a direct connection, escalating automatically to a shared datacenter proxy and then a residential one if a listing does not come through — the same connection is then reused for the rest of the run once it proves trusted. Responses are checked against known block and CAPTCHA signatures before being parsed, and a listing that eBay has genuinely removed is detected and reported quickly instead of retried forever.

Does it label feedback by whether it's about this listing or the seller overall?

Yes — every feedback row carries reviewScope, either this_item (buyers of this exact listing) or seller_all_items (the seller's feedback across all of their listings). If a listing has no resolvable seller login name, feedback lookup is skipped for that row and only the listing's own feedback counters are returned.

How many feedback entries does it return per listing?

Up to maxReviewsPerItem (default 20, maximum 500). Setting it to 0 asks for as many as eBay will serve, but the actor never returns more than 500 feedback entries for a single listing regardless — that ceiling is fixed in the code, not adjustable from the input. Listing/product rows are never limited by this setting.

How do I monitor a seller's feedback or a listing's price over time?

Save the listing URLs once, run the actor on a recurring Apify Schedule, and diff each new run's dataset against the previous one on product_price, product_condition, itemFeedbackCount, sellerFeedbackCount and the set of reviewComment values — alert when a new negative ratingType entry appears or a tracked field changes.

Does it work with Claude, ChatGPT or other AI agent frameworks?

It is not exposed through an MCP server, but it is callable as a standard HTTP endpoint through the Apify API from any agent framework that can make a request — an agent supplies listing URLs, receives the structured JSON rows back, and grounds its answer in real listing and feedback text instead of guessing.

Can I use it without managing proxies or an eBay account?

Yes. No eBay login is used anywhere in the run — only public listing and feedback pages are read. Apify Proxy is on by default and the actor escalates connection tiers by itself; the only thing you supply is the listing URLs.

Is my input from the original eBay Product Details Scraper still valid?

Yes. urls and startUrls are kept as legacy aliases for the same listing list — an older input keeps running unchanged and returns every field it used to, plus the fields documented here, as long as reviewSourceUrls (checked first) is left empty.

What happens if a listing has been removed or no longer exists?

It's detected and reported quickly rather than retried indefinitely: the row for that URL comes back with success: false and an error message such as "Listing not found or no longer available," so a dead link doesn't stall the rest of the run.

💬 Your feedback

Found a bug, or a field that doesn't match what's on the listing page? Open an issue from the Issues tab on this Actor's Apify Console page — the team behind eBay Product Details Scraper With Buyer Reviews & Ratings reads every one. Include the listing URL that misbehaved and it's usually enough to reproduce.