iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Start with the review platform’s official API, export, or licensed feed—not a browser scraper. Check that the source permits your intended collection and use, then verify what data it actually returns, how it may be stored or displayed, and how complete it is. Scraping rules differ by platform: Yelp says it does not allow scraping or copying its site content with third-party software, while Google Maps Platform terms prohibit scraping or exporting Maps content for use outside its services, including saving reviews. An API is not automatically permission to retain or republish everything it returns.
Choose an authorized way to collect the data
First define what you need: review text, ratings, product questions and answers, a date range, particular businesses or products, or aggregate themes. Then choose the narrowest permitted source that covers that need. The word “scraping” describes a collection technique; it does not establish permission.
| Route | Useful when | Check before implementation |
|---|---|---|
| Official platform API or export | The documented fields and coverage fit your project. | Eligibility, fields, quotas, pagination, regions, refresh cadence, attribution, retention, and permitted reuse. |
| Licensed feed or partner access | You need broader coverage or have a commercial data use case. | Which sources and records are included, and whether the license allows storage, combination, display, redistribution, or model training. |
| Direct web crawling | The site permits automated collection and no suitable authorized API or feed meets the need. | Current terms, robots.txt, request limits, identification, privacy, copyright, database, and jurisdictional requirements. |
These routes are not interchangeable. Yelp documents a review API endpoint but separately says scraping or copying its site content with third-party software is not allowed. Google Places policies govern attribution and storage for API content, while Google Maps Platform terms prohibit scraping Maps content for use outside its services. Amazon’s Customer Feedback API provides review-topic insights rather than an unrestricted dump of review text. See Yelp Places API documentation, Yelp’s scraping guidance, Amazon’s Customer Feedback API documentation, Google Places policies, and the Google Maps Platform terms archived June 4, 2025.
What official review APIs actually provide
Yelp Places API
Yelp’s documented reviews endpoint returns up to three review excerpts per business. That can help with a narrow display or a small amount of context, but it is not a complete business review archive. Do not represent those excerpts as all reviews or as a representative sample without evidence. Confirm the current endpoint fields, access conditions, and use terms in the Yelp Places API documentation.
Popular on Amazon
Amazon Customer Feedback API
Amazon describes this API for sellers and vendors. The documented operation exposes review-topic insights at ASIN and browse-node level; it is not an unrestricted review-text export. The documentation lists the US, UK, France, Italy, Germany, Spain, and Japan, says data is refreshed weekly and available only in English, and identifies a Brand Analytics role requirement for the documented operation. Those details describe the documentation reviewed in 2026; confirm that your account, marketplace, and intended operation qualify before building around it. See Amazon’s API documentation.
Google Places API
Using an API does not remove display and storage obligations. Google Places policies require review attribution and direct access to the source reviews, specify review-display requirements, and restrict caching or storage subject to stated exceptions. Check the current policy for the exact content and use case before saving or presenting returned reviews: Google Places API policies.
Check permission before writing a crawler
- Read the current terms and API or license rules. Determine whether your purpose includes collection, analysis, storage, display, redistribution, or commercial use. These permissions can differ. Yelp’s stated position and Google’s Maps terms are concrete examples of platform-specific restrictions.
- Inspect robots.txt, but do not treat it as a license. RFC 9309 describes a protocol by which site owners request that crawlers honor rules. It does not grant permission or replace the site’s terms and other applicable requirements. Read the IETF Robots Exclusion Protocol standard alongside the site’s terms.
- Assess the relevant jurisdictions and data. The target site, collector’s location, reviewers’ locations, purpose, personal information, and downstream use can all affect the analysis. This guide cannot determine whether a particular project is lawful; seek qualified advice if the consequences or uncertainty warrant it.
- Stop if access is denied or the site’s rules do not permit the collection. Do not try to get around a CAPTCHA, access control, or rate limit. Look for an authorized API, export, or licensed provider instead.
Build a reliable collection workflow
1. Define the data boundary
List the products or businesses, locales, date range, and required fields before collecting. Avoid reviewer identifiers or other personal information unless it is necessary and you have an appropriate basis and retention plan. A narrower specification reduces unnecessary collection and makes coverage easier to assess.
2. Record access conditions and expected coverage
For the chosen source, document who may access it, which markets and languages it covers, available fields, quotas, pagination, refresh interval, and use restrictions. For each query, note the expected result limit and whether records are excerpts, selected by ranking, or aggregates. A constrained endpoint cannot support a claim of completeness.
Also popular on Amazon
3. Collect conservatively
Use documented endpoints where available. If crawling is permitted, honor robots directives, identify the client where appropriate, use modest request rates, and apply retries with backoff for transient failures. Stop on access-denied or rate-limit responses rather than increasing request pressure. Keep a collection log with request time, source URL or API identifier, locale, query parameters, and API or schema version.
4. Preserve provenance while normalizing
Keep source IDs, product or business IDs, review or question IDs where supplied, timestamps, rating scale, language, and raw text separate from cleaned fields. Record translations and transformations instead of silently replacing the original. Deduplicate using stable identifiers where available; log edits or removals when the source exposes them. Do not assume that two similar texts are duplicates if their IDs or context differ.
5. Measure gaps and selection bias
Record which pages or API pages were fetched, failures, missing records, ranking or selection rules, and any result cap. Compare your collected counts with source totals when those totals are available. For analysis, stratify by date, locale, language, rating, and product or business variant when relevant. A handful of excerpts—such as Yelp’s documented maximum of three per business—cannot establish what all reviewers think.
Free tools Windows power users keep installed
One-click scans. No signup required.
6. Apply retention, attribution, and publication rules
Separate internal analysis from displaying or redistributing source content; the allowed uses may differ. Apply required attribution and source links, storage limits, and display context to both raw and derived data. Google Places, for example, requires attribution and direct source access for reviews and restricts caching or storage except under stated exceptions. Follow the source’s current policy rather than assuming that API access grants indefinite retention.
Also popular on Amazon
Handle review integrity in analysis and publication
Keep the source’s meaning intact. The FTC’s platform guidance says, “Don’t edit reviews to alter the message. For example, don’t change words to make a negative review sound more positive.” It also recommends reasonable processes to verify authenticity and equal treatment of positive and negative reviews. For projects that publish or feature reviews, explain selection criteria and do not present a selectively favorable set as representative. Read the FTC guide for platforms.
The FTC’s marketer guidance advises understanding the rules of the platforms where reviews appear and not soliciting reviews only from people expected to be positive. Its staff Q&A says the Consumer Reviews and Testimonials Rule took effect October 21, 2024, and cautions that the guidance is not definitive or comprehensive; context matters. See the FTC marketer guide and FTC Rule Q&A.
Amazon’s community guidance says to post only your own content or content you have permission to use on Amazon. Its promotional-content guidance says a person with a financial or close personal connection may answer product questions if that connection is clearly and conspicuously disclosed. These are examples of rules for contributing or displaying content on Amazon; they do not by themselves establish permission to scrape Amazon. See Amazon Community Guidelines and About Promotional Content.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesCompare collection routes on the facts that matter
- Permission: Does the source allow the intended collection and each downstream use?
- Coverage: Is the data full text or excerpts; all records or selected ones; reviews only or questions and answers too; which products, businesses, and locales?
- Freshness: How often does it update, and does it expose edits, deletions, or historical snapshots?
- Data quality: Are IDs, timestamps, rating scales, languages, verified-purchase labels, pagination, and provenance available?
- Access: Are there role or account requirements, geographic limits, quotas, prices, or authentication conditions?
- Retention and presentation: What storage duration, attribution, source links, display context, and combination or republication limits apply?
Document the answers with the date checked. For example, the Amazon API documentation identifies version 2024-06-01, listed markets, role requirements, and weekly refresh; Yelp documents a maximum of three review excerpts per business; and Google Places policies specify attribution and storage obligations. These are statements in the referenced documentation, not guarantees that future terms or API responses will remain unchanged.
Also popular on Amazon
Or skip the browser setup
If your goal is a screenshot of a review or Q&A page rather than a structured dataset, ScreenshotNeo is a website screenshot API and MCP server for developers. It captures PNG, JPEG, WebP, or PDF output; it does not turn reviews into a structured export or grant permission to collect content.
One GET request captures a page. See the ScreenshotNeo documentation.
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webp
ScreenshotNeo accepts cookie or consent banners before capture and removes 60+ known consent platforms, newsletter popups, and chat widgets; each step can be turned off. Bot checks, blank pages, timeouts, failed loads, and cache hits cost nothing, and responses identify the page verdict and billing status in headers. Its MCP server provides screenshot and PDF tools for AI agents. The free plan includes 1,000 screenshots a month with no card; paid plans start at $5 for 3,000.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Sign up free for 1,000 screenshots a month, with no card required.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Troubleshooting common collection problems
The API returns fewer reviews than expected
Check the documented cap, pagination behavior, filters, role eligibility, locale, and endpoint semantics. An endpoint returning excerpts or insights is not a full archive. Record its limit and describe your dataset accordingly rather than trying to bypass it.
Also popular on Amazon
Pages or API calls are missing records
Check for pagination tokens, query boundaries, transient failures, locale mismatches, and source-side ranking or filtering. Log every page or cursor and collection failure. Compare fetched counts against source totals if they are provided; if not, state that completeness could not be verified.
You see duplicates or conflicting versions
Prefer stable source IDs and preserve timestamps and source references. The same review can appear in multiple queries or pages; deduplicate on an identifier without discarding provenance. If the source changes a review, keep a version or change log only where your retention terms permit.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallYour crawler receives a denial or rate-limit response
Stop automated requests to that route. Recheck permissions and documented limits, then use an authorized endpoint or contact the provider about access. Do not rotate identities or evade controls to continue collection.
You cannot store or display API results as planned
Review the API’s separate storage, attribution, and display terms; access to retrieve content does not settle those questions. For Google Places reviews, check the required attribution, direct source access, display rules, and stated storage exceptions before implementation.
Frequently asked questions
Does robots.txt mean I have permission to scrape?
No. RFC 9309 describes crawler instructions; it is not a license. Check the site’s terms, applicable requirements, and the specific use you plan.
Can I use review data to train a model?
Do not assume so. Confirm that the API terms, source license, and applicable rules permit that use, including retention and any use of personal information. A general API credential is not proof of model-training rights.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →How should I describe a partial collection?
State the source, collection dates, locales, filters, endpoint limits, and known omissions. Distinguish excerpts or selected results from a complete dataset, and disclose when completeness could not be checked.

