Detecting inauthentic review trends across a sector

Discuss

Single fake reviews are noise. Coordinated inauthentic behaviour is signal. This article explains how sector-wide trend detection differs from per-review moderation, why broker networks operate at the category level, and how to read the public trend dashboard ScoreReview UK publishes each quarter.

From per-review to per-network detection

Per-review detection asks: is this one review fake? Network detection asks: are these 400 reviews part of the same operation? The maths is different — graph clustering, temporal correlation and embedding similarity replace simple text classifiers.

Sectors most exposed in 2026

Our quarterly enforcement log shows persistent campaign activity across moving services, cosmetic clinics, debt-advice firms and short-let property managers. These are sectors with high transaction value and weak repeat-purchase rates — exactly the conditions where a fabricated reputation pays back.

Temporal signatures of a campaign

Genuine reviews diffuse through a Poisson-like distribution. Brokered campaigns spike on weekday evenings UTC, cluster around payday windows, and stop abruptly when the broker is paid out. Plotting submission time against business age makes these patterns visible to the eye.

Cross-platform leakage

Brokers rarely buy from one platform. When we see a network on ScoreReview UK we can almost always find the same handles on three or four other consumer sites. We publish the deduplicated reviewer-handle list to participating platforms under a data-sharing memorandum.

What changes for honest businesses in a flagged sector

If your sector is under active enforcement, expect tighter invitation throttles, mandatory identity verification at Tier 4 for first reviews, and an extra 24-hour publication delay on unverified submissions. These measures protect the score you have already earned.

Reading the public enforcement log

Each quarter we publish action counts by category, the false-positive rate, the median time-to-action and the share of disputed decisions overturned. The log is the single best test of whether a platform's moderation is performative or real.

What regulators are watching

The CMA's 2025 review-platform compliance assessment singled out sector-level pattern analysis as a baseline expectation. Platforms that do per-review moderation only are now considered to be undershooting.

FAQ

Can I see the enforcement log?

Yes — it is public, versioned and downloadable as CSV. Every action carries an anonymised case ID for traceability.

What if my business is wrongly flagged in a campaign?

The dispute workflow gives you 72 hours to submit evidence; a senior moderator (not the original reviewer) makes the final call.

Keep reading

Get the newsroom in your inbox every Friday.

Reviews reporting, scam alerts and playbooks. Free. Unsubscribe anytime.

Prefer a demo? Talk to us

Trending#DMCC#CMA#compliance#UK law#moderation#GDPR#data protection#retention#AI#reply automation#brand voice#sentiment#analytics#hospitality

Discussion (2)

Priya S.· 2 days ago
Really practical breakdown — the four-part reply structure is now on our till-side crib sheet. Thank you.
Dan (Cannock Plumbing)· 5 days ago
Went from 12 reviews to 47 in three months following almost exactly this playbook. It works.