Local Business Review Spike & Rating Monitor

Buyer task: Agencies and multi-branch businesses audit the freshness of aggregate review data and unexpected changes in rating trajectories.

Detect branch-level review volume per day spikes/drops and aggregate rating declines in user-provided review-count exports, without collecting personal reviews.

Who this is for

Teams managing repeated vendor, catalog, branch or listing data exports who need an audit trail of historical changes and data-quality anomalies. Treat this output as decision support for business operations, not a guarantee of rankings, marketing conversions or revenue.

How to use

1. Start with the fully fictional demonstration inputs already provided in the Actor input form.
2. To analyze your business, replace both JSON array fields with two compatible dated exports that you already own or are permitted to analyze. The two inputs are called "previous" and "current".
3. Keep business identifiers stable (ASIN and seller ID, store and SKU, or business location ID as appropriate). If the original sources have inconsistent identity keys, normalize them before comparing.
4. Execute a bounded run, then download the findings from the default dataset as JSON/CSV. See the SUMMARY key in the default key-value store for the total number of findings and high-priority records.
5. Investigate high-severity issues first, but confirm changes against your authoritative system before any commercial or storefront modification.

Why compare over time

Agencies and multi-branch businesses audit the freshness of aggregate review data and unexpected changes in rating trajectories. A weekly or daily analyst can use fresh exports with an unchanged schema to monitor recurring risks. This Actor is deliberately stateless: it does not fetch new platform data, automatically schedule data collection, or retain previous snapshots for a future run. Customer-supplied rows are not mixed with other accounts.

Output fields

recordType, entity, issue, severity, before, after, details, sampleWarning. Each row explains what changed or why comparison is not reliable. Inputs are capped at 1,000 records in each snapshot; repeated or invalid identifiers are rejected instead of silently mixed.

Actual billing and safety

A report pay-per-event charge occurs only after a successful dataset and summary save. Consult the live Pricing tab for the current USD event price and possible actor-start charge. Viewing the public fictional example does not require running this Actor. No paid external API, browser bypass, private login, proxy, user contact information or third-party scraping is used.

Evidence limitations

Only fields provided in imported data are analyzed. Missing observations are NOT proof of actual product deletion, shop closure or seller delisting. Cross-currency prices are never converted without trustworthy declared FX evidence. Review scores are aggregate indicators and not causal evidence; never send personal reviewer names or contact information. Listed prices are not sales, and product offer changes do not show actual stock counts or transactions.