Microsoft Marketplace Reviews Scraper
Pricing
from $0.50 / 1,000 saved reviews
Microsoft Marketplace Reviews Scraper
Unofficial exporter for public Microsoft Marketplace and AppSource reviews, ratings, dates, sources, and product metadata as structured JSON.
Pricing
from $0.50 / 1,000 saved reviews
Rating
0.0
(0)
Developer
Đăng Phạm
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
8 days ago
Last modified
Categories
Share
Export public Microsoft Marketplace and AppSource ratings and review text into clean, structured records for competitor research, product-feedback analysis, vendor monitoring, voice-of-customer workflows, and downstream API or agent automation.
This independent community Actor is not affiliated with, endorsed by, or sponsored by Microsoft. It uses public Marketplace storefront data and does not require a Microsoft login.
Supported product URLs
https://marketplace.microsoft.com/.../product/...https://appsource.microsoft.com/.../product/...https://azuremarketplace.microsoft.com/.../marketplace/apps/...
Only HTTPS product pages on those exact domains are accepted. Arbitrary crawl URLs, private hosts, reviewer profiles, and credentialed endpoints are rejected.
Input
| Field | Type | Default | Description |
|---|---|---|---|
productUrls | string[] | required | 1–20 supported product URLs |
maxReviewsPerProduct | integer | 100 | 1–1,000 review rows per product |
ratingFilter | integer[] | all | Any combination of 1–5 stars |
sortOrder | string | newest | newest, oldest, highest_rating, lowest_rating, or most_helpful |
locale | string | en-us | Optional language-region override |
includeProductSummary | boolean | true | Emit a summary row for each product; empty results always receive one |
Example:
{"productUrls": ["https://marketplace.microsoft.com/en-us/product/office/WA200004820"],"maxReviewsPerProduct": 25,"ratingFilter": [1, 2, 3],"sortOrder": "most_helpful","locale": "en-us","includeProductSummary": true}
Output
The default dataset contains flat review rows and deterministic product_summary rows. Fields are emitted only when the public source exposes them.
{"recordType": "review","productId": "WA200000001","productName": "Example product","publisherName": "Example publisher","productUrl": "https://marketplace.microsoft.com/en-us/product/office/WA200000001","aggregateRating": 4.125,"aggregateRatingCount": 48,"reviewId": "public-review-id","reviewSource": "microsoft_marketplace","reviewRating": 5,"reviewTitle": "Useful integration","reviewText": "Public review text appears here.","reviewDate": "2026-08-01T12:00:00Z","helpfulCount": 3,"locale": "en-us","scrapedAt": "2026-08-24T12:00:00.000Z"}
reviewSource is one of microsoft_marketplace, g2, or other_public_source. AppSource, Teams, Marketplace, and Ibiza storefront sources are grouped as microsoft_marketplace; G2 remains distinct. Unknown public source labels are not guessed and map to other_public_source.
For a product with no matching review text, the Actor saves a product_summary row with aggregateRatingCount and reviewRecordCount rather than failing the run.
API and MCP
Run through the Apify API after deployment:
curl -X POST "https://api.apify.com/v2/acts/USERNAME~microsoft-marketplace-reviews-scraper/runs?token=YOUR_APIFY_TOKEN" \-H "Content-Type: application/json" \-d @sample-input.json
Never commit or share the token. Once the Actor is public, the standard Apify API integrations and Apify MCP Server can expose the Actor to compatible automation and agent clients; no custom MCP server or Microsoft credential is required.
Pricing
The live Store model is pay per event with platform usage included:
start: USD 0.005 once, charged only after the first review is successfully saved.review-result: approximately USD 0.0005 per successfully saved review (about USD 0.50 per 1,000 reviews), charged after storage succeeds.
Product summaries, empty products, failed requests, retries, duplicate rows, skipped rows, and parser errors are not billable events. The Store page displays USD 0.005 for the one-time start event and USD 0.50 per 1,000 review-result events.
Public data and privacy
The Actor collects only public Marketplace product and review fields. It does not enrich reviewers, crawl reviewer profiles, collect email/phone/address data, attempt de-anonymization, or combine reviews with external identity sources. Anonymous reviews remain anonymous.
Reliability and limitations
- Microsoft documents Marketplace ratings/reviews and G2 content, but the unauthenticated review host used by the storefront is an undocumented public dependency. It is not presented as an official supported Microsoft API and can change without notice.
- Schema validation, request timeouts, bounded exponential retry, source-host allowlists, response limits, duplicate suppression, and partial-product isolation reduce failure impact.
- The public storefront currently returns a bounded review array and can cap large products. The Actor does not guarantee complete review history.
- Ratings without title or text contribute to the aggregate but do not become review rows.
- Availability, regional differences, removed reviews, and storefront changes can affect results.
- This is a data export tool, not legal advice. Users are responsible for lawful use and applicable terms.
Microsoft's public documentation: Ratings and reviews. Apify conventions used by this Actor: Actor definition, dataset schema, and pay-per-event charging.
Troubleshooting
- Unsupported URL: pass a product detail URL on one of the three exact supported domains.
- Product metadata schema error: the public catalog response changed or the listing is unavailable; retry later and report the product URL.
- Review schema error: Microsoft changed its storefront response. Other products still complete; inspect the run log for the failed URL.
- No reviews: a product summary is expected and is not an error.
- Partial run: one broken product is isolated so successful products remain in the dataset.
Support
Open an Actor issue with the product URL, run ID, timestamp, and sanitized error message. Do not include Apify tokens, Microsoft credentials, private browser data, or complete third-party review bodies.