Amazon Review Analyzer — ReviewMeta Alternative

Analyze an Amazon review export for repeated text, similar long reviews, and concentrated rating-date patterns. Get a duplicate-aware sample average and transparent flags from the review records you supply. For product researchers, the dataset returns one aggregate pattern report for the supplied Amazon review sample.

Workflow: put the results to work

Supply authorized review rows directly or through a dataset ID. Inspect duplicate groups and rating-date concentrations before interpreting the adjusted sample average. These are patterns within your supplied sample, not a verdict that a reviewer or product is fraudulent.

What it does

• Reads an inline JSON review array or an accessible Apify dataset containing an authorized review export.
• Understands common Amazon-review field names such as review_id, rating, review_body, review_date_iso, and is_verified_purchase.
• Produces exactly one homogeneous review-pattern-analysis row with source provenance, sample coverage, rating counts, pattern signals, and method thresholds.
• Calculates a displayed sample average and an exact-duplicate-aware average. An exact text cluster contributes at most one full rating weight.
• Writes COMPLETE, PARTIAL, VALID_EMPTY, INVALID_INPUT, UPSTREAM_FAILED, or CONFIG_ERROR terminal outcomes to both OUTPUT and RUN_SUMMARY.

What it does not do

• It does not fetch Amazon product or review pages.
• It does not identify a person, infer intent, or determine that any review is fake, genuine, incentivized, or deceptive.
• It does not replace a complete customer-review platform, browser extension, price research, seller monitoring, or historical review archive.
• It does not write ratings, contact reviewers, change listings, or retain review text in the output row.

Choose the right tool

Use this Actor when you already have a review export and need a reproducible, source-linked pattern report in JSON. Choose ReviewMeta where its existing hosted consumer experience, historical analysis, extension, or proprietary workflow is available and better fits the decision.

For collection, first obtain reviews through a source you are authorized to use. Pass that export directly in reviews, or provide its accessible Apify sourceDatasetId. The Actor intentionally keeps collection separate from assessment so the evidence boundary stays clear.

Workflow story

An analyst starts with a permitted review export for one product. They then submit the rows or the export dataset ID, compare the displayed and duplicate-aware averages, and inspect the aggregate signals with the source coverage. Next, they can export the one report row or feed it into their own review workflow alongside the original authorized source.

How this workflow compares with ReviewMeta

Decision	This Actor	ReviewMeta	Best fit
Scope and workflow	Aggregate analysis of caller-supplied review exports	Hosted consumer review-analysis experience	Select the scope that matches the workflow.
Billing and cost	Pay per event for a stored report plus platform usage	Consumer-facing offering with no matched automation allocation in the reviewed source	Review current pricing for the chosen route.
Same-job cost	A bounded report event is visible in the run	No like-for-like automation allocation documented	Treat cost as a documented tradeoff.
Effective efficiency	One report row carries method, provenance, and aggregate signals	Hosted consumer workflow	Use structured output when it fits the operating flow.
Integrations and automation	Apify API, datasets, schedules, and webhooks can compose the report workflow	Hosted product experience	Choose the integration surface that is required.

ReviewMeta is a trademark of its owner. This independent Actor is not affiliated, associated, or endorsed by ReviewMeta.

Input

Provide exactly one review source:

1. reviews: a JSON array of review objects, or
2. sourceDatasetId: an accessible Apify dataset ID containing review rows.

Every valid review requires a numerical rating from 1 to 5. These aliases are accepted:

Meaning	Accepted fields
ID	review_id, reviewId, id, idReview
Rating	rating, star_rating, starRating, ratingScore
Text	review_body, reviewDescription, body, text, content
Title	review_title, reviewTitle, title
Date	review_date_iso, reviewedAt, reviewDate, date, review_date
Verified purchase	is_verified_purchase, isVerified, verifiedPurchase, verified

At least three valid rows are needed to create a report. A pattern read is labeled INSUFFICIENT_SAMPLE if it has fewer than ten valid reviews or fewer than six dated reviews.

{
  "productAsin": "B07DF46NW9",
  "productUrl": "https://www.amazon.com/dp/B07DF46NW9",
  "maxReviews": 200,
  "reviews": [
    {
      "review_id": "R1",
      "rating": 5,
      "review_title": "Useful for daily use",
      "review_body": "I used this for several weeks and found the setup simple.",
      "review_date_iso": "2026-01-15T00:00:00.000Z",
      "is_verified_purchase": true
    },
    {
      "review_id": "R2",
      "rating": 4,
      "review_body": "Good overall, although delivery took longer than expected.",
      "review_date_iso": "2026-01-18T00:00:00.000Z",
      "is_verified_purchase": true
    },
    {
      "review_id": "R3",
      "rating": 5,
      "review_body": "The item matched the listing and has held up so far.",
      "review_date_iso": "2026-01-20T00:00:00.000Z",
      "is_verified_purchase": true
    }
  ]
}

To analyze an existing dataset instead, omit reviews:

{
  "sourceDatasetId": "ACCESSIBLE_DATASET_ID",
  "productAsin": "B07DF46NW9",
  "maxReviews": 200
}

Output

One result means one aggregate report—not one review. The dataset output never reproduces reviewer names, profile URLs, or review text.

{
  "reportType": "review-pattern-analysis",
  "product": { "asin": "B07DF46NW9", "sourceUrl": "https://www.amazon.com/dp/B07DF46NW9" },
  "sample": { "validReviewCount": 120, "datedReviewCount": 114 },
  "ratings": {
    "displayedAverage": 4.42,
    "duplicateAwareAverage": 4.31,
    "ratingCounts": { "1": 8, "2": 6, "3": 10, "4": 32, "5": 64 }
  },
  "assessment": {
    "label": "REVIEW_PATTERN_CHECK",
    "confidence": "medium",
    "explanation": "The sample contains patterns worth reading in context before relying on the displayed average."
  },
  "signals": [
    {
      "code": "EXACT_TEXT_DUPLICATION",
      "level": "review",
      "title": "Repeated normalized review text"
    }
  ]
}

REVIEW_PATTERN_CHECK means the configured sample pattern deserves contextual review. It is never a conclusion that reviews were manipulated. NO_STRONG_PATTERN_IN_SAMPLE only means the configured checks did not find a signal in the supplied rows; it is not an authenticity guarantee.

Method and limitations

• Exact duplication: lowercases and Unicode-normalizes title/body, removes punctuation and repeated spaces, then compares text at least 24 characters long. Only these exact clusters affect the duplicate-aware average.
• Near similarity: reports only long-text pairs of at least 100 characters with token-set Jaccard similarity at least 0.90. These pairs never change the rating.
• Date concentration: reports a rolling seven-day window containing at least six dated reviews where 80% or more share the same numerical rating. A campaign, launch, seasonality, or an ordinary short sample can create this signal.
• Unseen evidence: deleted reviews, seller history, marketplace controls, purchases not present in the export, reviewer networks, and product-quality evidence are outside the report.

Use the report as one reason to inspect evidence, not as a purchasing, moderation, regulatory, employment, credit, insurance, or eligibility decision by itself.

API and agent prompt

curl "https://api.apify.com/v2/acts/khadinakbar~reviewmeta-alternative/runs?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d @input.json

Use with an AI agent through Apify MCP

Analyze this authorized Amazon review export. Return a duplicate-aware rating, sample coverage, pattern signals, and the method limitations. Do not call any review fake and do not infer reviewer intent.

After the run, have the AI agent read the outcome and dataset report, preserve the supplied source provenance, and present the billed report cost and scope alongside the pattern assessment.

Best results

• Use a consistent export from one product and marketplace per run.
• Prefer a substantial review sample with parseable dates for a more useful pattern comparison.
• Preserve original review IDs and dates in the source data; they improve deduplication and date coverage.
• Keep the raw export in your permitted source dataset. This Actor returns only aggregate evidence, so source-level context remains auditable outside the report.

Builder's note

I designed this Actor around transparent restraint. Review-pattern signals are practical when their thresholds, sample boundary, and blind spots travel with the result. This design keeps the aggregate output useful while preserving the original review evidence for human context.

Responsible use

Analyze data you are authorized to access and collect, follow applicable laws and source terms, and treat the report as a structured input to a contextual review process.

Pricing and run costs

This Actor uses Pay per event plus Apify platform usage. The Pricing tab lists the current event rates and billing terms.

Event	Billing unit	When it applies
apify-actor-start	Actor Start	Charged when the Actor starts running. Number of events charged depends on Actor memory (one event per GB, minimum one event).
review-pattern-report	Review Pattern Report	Charged once when one validated aggregate review-pattern report is stored. Invalid or too-small inputs do not create this event.

Run cost combines the charged events and Apify platform usage. Review the run charge limit and requested result count before starting.

Connect an AI agent

Use the Apify MCP configurator to choose an available client connection. Inspect this Actor’s current input schema and required credentials before running it.