The wall
Every review across both captures, one cell each, oldest to newest. Green cells are reviews classified likely genuine — almost all of them detailed complaints. Red cells carry hard fabrication markers. Gray cells are 5-star reviews that cannot be verified either way, overwhelmingly posted by accounts with one or two lifetime reviews — and note how many gray cells turned red as fuller review text surfaced more template families. Hover or tap any cell for detail. This is what a burial looks like.
Two populations, one listing
Compare the accounts behind the 5-star wall with the accounts behind the complaints. If both were real customers of the same company, these distributions should look alike. They don't.
Accounts with ≤2 lifetime reviews
223 of 279 five-star reviews come from accounts that have reviewed almost nothing else on Google, ever.
Google Local Guides (established reviewers)
Local Guides — accounts with long, verifiable review histories — are five times more common among the complainants.
Reviews the owner replied to
The owner warmly engages the praise — often restating its exact details back — and largely ignores the documented complaints.
Exhibits A–H
Eight independent categories of evidence, each visible in the public record without any access to private data. Individual reviews can't be forensically proven fake from the outside; patterns can be documented. Click an exhibit to expand it.
A real customer gamed the sort order to warn you. A Google Local Guide posted a scathing warning — and deliberately rated it 5 stars so it wouldn't be buried beneath the wall of praise. That a customer felt this was necessary is itself evidence of how the listing behaves.
The single most conclusive artifact class. The same idiosyncratic story — the antique mirror, the grand piano the movers "tuned," the aquarium, the thunderstorm, grandma's china cabinet, the daughter's first apartment — retold under multiple unrelated account names, sometimes verbatim, sometimes lightly paraphrased. Real customers do not share these coincidences; outputs of the same writing prompt do.
Three more accounts — Phillip West ("wrapped that mirror like it was glass from a museum"), Darryl Brown ("wrapped it like belonged in a museum. Price stayed exactly what…"), and Victor Bailey — retell the same mirror story in paraphrase. Five accounts, one mirror.
Also documented: the thunderstorm/flooded-streets pair (Pierce Hamm / Donna Mitchell), the aquarium "my fish survived" pair (Bill Howard / Daniel Scott), the china-cabinet pair (Eula Bickford / Billy Bell), the daughter's-first-apartment pair (Lawrence Peterson / Carl Rivera), the route-swap pair ("Alabama to Pompano" / "Minnesota to Boca" — otherwise the same review, Kole Whitten / Jason Caldwell), the verbatim "AaSII ASIT" / Jared Conway pair, and a shared stylistic seed that closes unrelated reviews with a sunset ("The sunset that day/evening was incredible/beautiful" — 4 accounts) or a philosophical shrug ("moving feels weird after… life just moves on I guess" — 4 accounts).
Nine "customers" moving professional equipment from West Virginia to South Florida. The template is "[WV town] to [FL city] with my [profession] equipment… price matched/reasonable": dental equipment from Williamson (Steve Roberts), gym equipment from Huntington (David Kelly), art supplies from Harpers Ferry (Frances Allen), nursing equipment from West Virginia (Story Fryar), a tool chest from "that tiny stretch by the Tug Fork" (Colby Mehta), farm equipment from "rural West Virginia" (Donna Mitchell #1), plus Weirton-to-Miami twice (Aaron Griffin / Nancy Harris #1 — near-verbatim of each other, both featuring "Alex" who "didn't talk much") and a generic WV move (Billy Turner). A Fort Lauderdale broker whose 5-star reviewers are, at implausible rates, West Virginians with equipment — because the generator was seeded with a West Virginia town list.
Verified against the @faker-js/faker name corpus — the same fake-data library documented in the Facebook audit. Google reviewers whose names come straight from the corpus: Shany Dare and Alexandro Balistreri (first and last names are both corpus entries), Boyd Bauch, Valerie Stiedemann, Teresa Mante, Eula Bickford. Balistreri, Stiedemann, Bauch, and Mante are vanishingly rare real-world surnames; the odds of several organically reviewing one Florida mover are effectively zero. The Facebook cohort drew from the identical corpus (Hegmann, Haag, Emmerich, Homenick…).
Reused persona names across separate accounts: two different accounts named "Ralph Cooper" (each showing "1 review" — impossible for one account with two live reviews), two named "Nancy Harris," two named "Donna Mitchell" — each pair participating in different template families. The persona pool recycles its own parts, exactly as on Facebook.
Purchased-account artifacts: reviewer names that are Thai school-roster entries — "26 ณัฐริกา ไพบูลย์," "18 ด.ช ศิววงศ์ วงศ์แสงกุล," "24 ขวัญฤทัย หงษ์ใหญ่" (a roster number plus "ด.ช.", the Thai honorific for a schoolboy) — alongside Korean ("김유빈30205") and Taiwanese accounts, all with 1–2 lifetime reviews, all writing fluent-English 5-star reviews for a Florida moving broker. One of them slips into the third person: "The movers hauled her bed, desk, and boxes" — the text was generated about the persona, not by it. Plus the celebrity accounts: Leonardo DiCaprio, Samuel L. Jackson, and generator-corrupted "AaSII ASIT."
The praise is shaped like a rebuttal. With full review texts, 41 five-star reviews volunteer that the price didn't move: "matched the quote," "final bill matched," "the quote stayed what they said," "what they said is what I paid," "price stayed exactly as quoted," "within $50 of the quote." Real happy customers rarely volunteer that their bill matched their quote — but it is precisely the line you would script if the genuine complaints all alleged a moving-day bait-and-switch. The identical rebuttal ran at industrial frequency on Facebook ("no hidden fees, no surprise charges" — 25 exact repetitions).
The Facebook and Google campaigns draw from one story bank — proving a single operator behind both. The Facebook audit documented recycled families including "Moving day started with a thunderstorm…" (3 FB accounts), antique-mirror variants, aquarium variants, and "I wasn't gonna write a review, but after they handled my [item]…" On Google, the same seeds resurface under new names: the thunderstorm (Pierce Hamm / Donna Mitchell #2), the antique mirror (5 accounts, Exhibit B), the aquarium (Bill Howard / Daniel Scott), and "I wasn't going to leave a review, but after they handled my antique mirror…" (Nancy Harris / Charles White). Same generator, same seed list, two storefronts.
Five 5-star reviews in the 24 hours before capture — every one from an account with exactly two lifetime reviews. They landed at 52 minutes, 13 hours, 14 hours, and 22 hours (×2) before the snapshot, all flagged NEW. In the same hour as the last two, a genuine customer posted a detailed complaint about losing a ~$4,000 deposit — and was immediately sandwiched between two of the fresh 5-stars. Over the prior two weeks: 24 five-star arrivals against 6 complaints. On Facebook this same behavior was witnessed in real time and timestamped to the minute; on Google, the relative-date display only permits day resolution — and the pattern still shows.
Owner replies recite the review back, noun for noun. Replies to 5-star reviews mechanically restate the review's specific details — the anti-static wraps, the narrow driveway, the Andrews Ave parking — as if generated from the review text itself. The owner replied to 76% of 5-star reviews and 25% of complaints; replies to detailed complaints, when they exist, are boilerplate or claim no record of the customer.
Posting tempo, working backward from capture
Complaints arrive at a steady trickle, the way real dissatisfaction does. The 5-star volume surges — including a measurable acceleration in the final two weeks of the window.
What the complaints describe
The 88 likely-genuine reviews span more than two years, include Local Guides with hundreds of prior reviews, and don't know each other. Independently, they describe the same repeating sequence documented in the Facebook victim record: lowball quote, non-refundable deposit, moving-day re-quote, goods or deposit held. Counts are reviews in this capture whose text raises each allegation.
The full record — all 367 reviews, classified
Every review from the capture with its account signals and classification. Filter, search, and sort. If a glowing review of this company was posted by a name below with one lifetime review, weigh it accordingly. Classifications are graded assessments from documented markers — not statements of fact about any individual reviewer.
| Reviewer | Rating | Account | Classification | Review & basis |
|---|