Amazon Reviews Scraper + AI Product Research & Opportunity Finder

Scrape Amazon reviews and turn them into a seller-facing product-opportunity verdict — complaint clusters, unmet-need gaps, differentiation angles, and a go / no-go score — plus a shareable HTML report for every product. This Amazon review scraper does the analysis most tools leave to you: it reads the review corpus, finds the recurring pain points, and tells you whether there's room to build a better product. And it never charges you for an empty run — the number-one complaint about every other Amazon review scraper on the market.

Where a SaaS suite like SmartScout ($29–187/mo), Helium 10 ($39–279/mo), or SellerSprite ($70/mo) locks review analysis behind a monthly subscription, this actor runs pay-per-use: analyze one product for a few cents, or a hundred on a schedule, and pay only for verdicts that actually land.

How does the Amazon Review Analyzer work?

The actor collects Amazon review signals and distills them into a structured product-opportunity report:

• Complaint clusters — the recurring problems buyers report, each with a severity rating and rough mention share (frequency × severity is exactly how experienced sellers prioritize).
• Unmet needs & gaps — features and fixes customers keep asking for, i.e. demand you could capture.
• Differentiation angles — concrete ways a new entrant could beat the incumbent, grounded in the review evidence.
• What the incumbent owns — the praise pillars you'd have to match to compete.
• Opportunity verdict — a 0–100 score, a demand-vs-saturation read, and a plain-English go / conditional-go / no-go recommendation.
• Shareable HTML report — a clean, self-contained report page per product (public URL), ready to send to a client or partner.

Three ways to get the reviews (pick your depth)

Amazon gates its full review list behind a login, so this actor gives you three input modes instead of one brittle scraper that breaks on it:

1. Snapshot mode (anonymous, zero setup — recommended default). Scrapes the public product page: top reviews, the full star-rating histogram, and Amazon's own "Customers say" aspect sentiment analysis (per-aspect summaries with representative quotes). Rich, instant, no login, no cookies.
2. Cookies mode (full depth). Paginates the complete review list using your own Amazon session cookies — for when you need every review, not just the product-page sample.
3. Dataset mode (analyze what you already have). Point the actor at any existing Apify dataset of reviews — from any review scraper — and get the opportunity verdict on data you've already collected.

Amazon Review Analyzer vs. other tools

	This actor	junglee/amazon-reviews-scraper	SmartScout / Helium 10 / SellerSprite
Pricing model	Pay-per-use (from ~$0.02/product)	Pay-per-use	$29–279 / month subscription
Charges on empty runs	Never	Yes (top complaint)	N/A
Multi-product batch	Yes, reliable	Frequently broken	Yes
Product-opportunity verdict	Built in	No (raw reviews only)	Partial / add-on
Shareable HTML report	Yes	No	No
Analysis included, no key setup	Yes	N/A	N/A
Analyze an existing dataset	Yes	No	No

The incumbent Amazon review scraper on the Apify Store has over 10,000 users but a 2.34★ rating — because it charges for empty runs, caps reviews, and breaks on multi-product input. This actor is built to fix exactly those failures and add the analysis layer on top.

What data can you extract from Amazon reviews?

Every run outputs one structured row per product, containing:

• Product title, brand, ASIN, average rating, and total rating count
• The full star-rating histogram (percentage at each of the 5 star levels)
• Amazon's "Customers say" summary and per-aspect sentiment chips (with representative snippets)
• The scraped review corpus (in cookies mode) or product-page reviews (in snapshot mode)
• The complete insight object: complaint clusters, unmet needs, differentiation angles, praise pillars, and the opportunity verdict
• A public URL to the rendered HTML opportunity report

How to use the Amazon Review Analyzer

1. Choose a mode (start with snapshot — no setup).
2. Paste one or more ASINs or Amazon product URLs (multi-product batches work reliably).
3. Leave Generate verdict on to get the full opportunity report.
4. (Optional) Add competitor ASINs for side-by-side context, or switch to cookies/dataset mode for full-corpus depth.
5. Run it — each product returns a data row plus a shareable report link. Schedule it to track how a niche's complaints and ratings shift over time.

Pricing

Pay-per-event — you pay for results, never for empty runs.

Event	Price	When it fires
Product snapshot scraped	~$0.02 / product	Snapshot mode, per product with real data
Review scraped	$0.0025 / review	Cookies mode, per review collected
Product opportunity verdict	$0.40 / product	Per opportunity report generated
Analysis overflow	$0.20 / 1,000 reviews	Only above your analysis cap

Example: a snapshot + opportunity verdict on one product costs about $0.42 — versus a $29–$70/month subscription you'd pay whether you analyze one product or none. Research ten products for roughly the price of a coffee, with no monthly commitment.

Example input

{
  "mode": "snapshot",
  "asins": ["B09B8V1LZ3", "B0BSHF7WHW"],
  "marketplace": "com",
  "maxReviewsAnalyzed": 1000
}

Example output (abridged)

{
  "asin": "B09B8V1LZ3",
  "title": "Amazon Echo Dot (newest model)",
  "avgRating": 4.7,
  "totalRatings": 195514,
  "histogram": { "5star": 82, "4star": 11, "3star": 3, "2star": 1, "1star": 3 },
  "insight": {
    "complaintClusters": [
      { "label": "Connectivity drops", "severity": "high", "mentionShare": "~18%", "summary": "Repeated Wi-Fi disconnects requiring re-setup." }
    ],
    "unmetNeeds": ["Reliable multi-room sync", "Privacy-first mode with a hardware mic cut"],
    "differentiationAngles": ["Position on rock-solid connectivity vs. the incumbent's drop complaints"],
    "verdict": {
      "opportunityScore": 62,
      "recommendation": "conditional-go",
      "oneLine": "Enter only with a defensible wedge on connectivity or privacy."
    }
  },
  "reportUrl": "https://api.apify.com/v2/key-value-stores/.../records/report-B09B8V1LZ3"
}

Who uses this Amazon review scraper?

• Amazon FBA sellers doing product research — decide whether a niche is worth entering before you source inventory.
• Private-label brand builders — turn a competitor's 1- and 3-star reviews into your product roadmap.
• Product researchers & sourcing agents — screen dozens of ASINs a day for solvable, high-frequency complaints.
• Agencies & consultants — hand clients a branded, shareable opportunity report instead of a spreadsheet.
• Market analysts — track how sentiment and complaint themes shift across a category over time with scheduled runs.
• Existing scraper users — already have review data in Apify? Run dataset mode and get the verdict without re-scraping.

Frequently asked questions

How many Amazon reviews should I analyze for product research? Around 100 reviews is enough for a single product read, and 500+ across competitors reveals category-level patterns. 3-star reviews carry the most balanced signal. The actor's analysis cap defaults to 1,000 — plenty for theme convergence — and you're only billed for overflow beyond it.

Can it analyze all of a product's reviews? Amazon requires a login to page through the full review list. Snapshot mode returns the product-page reviews plus Amazon's own aspect-level analysis (rich, and enough for a solid verdict); cookies mode, using your own Amazon session, unlocks full review-list pagination.

Does it charge me if a product has no reviews or the run fails? No. Charging is guarded — an empty result produces no billable event. This is the single biggest fix versus the incumbent Amazon review scraper.

Do I need my own API key or a separate subscription? No. Everything is bundled — no bring-your-own-key setup and no extra accounts to manage. It works out of the box.

Can I compare two products side by side? Yes. Add competitor ASINs and each is scraped and analyzed alongside your primary product for a direct comparison.

Which Amazon marketplaces are supported? US (.com) is the primary target, with .co.uk, .de, .ca, .com.au, .fr, .es, and .it selectable.

Is this cheaper than Helium 10 or SmartScout for review analysis? For occasional or batch product research, yes — you pay a few cents per product instead of $29–$279 every month, with no subscription.

Can I analyze reviews I already scraped with another tool? Yes — dataset mode reads any Apify dataset of reviews and runs the opportunity verdict on it, no re-scraping needed.

Limitations (the honest version)

• Full review pagination requires login. Snapshot mode is anonymous and returns the product-page reviews plus Amazon's aspect analysis — genuinely rich, but not the entire review corpus. Use cookies mode with your own session for full depth. High-volume pagination on a personal account carries the usual Amazon account-activity risk; use an account you're comfortable with.
• Turning the verdict off gives you scrape-only output. With Generate verdict unchecked, the actor still scrapes and outputs the full review data — just without the opportunity report.
• Amazon changes its page structure frequently. The parser is built with fallbacks and degrades gracefully, but layout drift can occasionally reduce what's captured on a given day.
• Residential proxies are strongly recommended for Amazon and are used by default.