We analysed 1,373 restoration companies and read 8,685 homeowner reviews across 21 US metros. Star ratings turned out to be nearly useless — and complaints were about something other than the work.
87% of restoration firms are rated 4.5 or higher, so star ratings barely separate anyone. Busier firms rate slightly higher, not lower. And homeowners almost never criticise the restoration work itself — they criticise billing, estimates and insurance handling.
We built this directory by scraping every restoration company Google Maps would surface across 21 US metros, then reading the review text to write our city rankings. That left us with a dataset nobody in this industry seems to have looked at: 8,685 individual reviews, roughly 2.4 million characters of homeowners describing what actually happened when their house flooded.
Four things in it surprised us. One of them changes how you should read any restoration company's star rating.
43.7% of restoration firms have a perfect 5.0. 87% are rated 4.5 or above. Fewer than one in twenty sits below 4.0. The scale is so compressed at the top that a star rating barely separates anyone.
The reason is sample size. The median firm in our dataset has just 37 reviews, and 42% have fewer than 25. A perfect score is easy to hold when only a handful of people have voted.
Of the 544 companies showing a flawless 5.0, 335 of them got there on fewer than 25 ratings. That isn't evidence of excellence — it's evidence of a small sample. It's the single most misleading number a homeowner sees when comparing crews in an emergency, and it's why our own rankings weight rating against review volume rather than taking the average at face value.
We expected the opposite. In most service industries, ratings erode as volume grows — more jobs, more chances to disappoint, more unhappy customers motivated to write. Restoration runs the other way.
The gap is small but it runs consistently in the unexpected direction. The most plausible explanation is selection: restoration is bought under duress, usually on a recommendation or a search made at 2am, and firms that survive to hundreds of reviews in that environment are the ones that stopped generating bad ones. Volume here is a filter, not a liability.
Practically, that means the intuition “I'll pick the small local outfit with the perfect score over the big one with a 4.7” is backwards in this industry more often than not.
This is the finding we'd want any homeowner to take away. We separated the review text into praise (5-star) and criticism (1–3 star), normalised for the very different volumes, and compared which words appear where.
Criticism is overwhelmingly about money and paperwork. Praise is overwhelmingly about thoroughness and punctuality. Complaints about the quality of drying, extraction or repair barely register.
The single most complaint-skewed word in the entire corpus is refund, followed by charge, wait and delay. Meanwhile insurance shows up more than twice as often in criticism as in praise — the paperwork, not the pumps, is where these jobs go wrong. If you are vetting a crew, the evidence says to interrogate the estimate, the billing process and who handles your adjuster, rather than the equipment list.
Water damage is an emergency category — burst pipes don't wait for business hours. Yet only 55% of the firms we scraped publish 24-hour availability on their Google listing. Nearly half are, at least officially, a weekday business.
There's a related pattern in how firms engage: 57% of the reviews we sampled had an owner response. Responding to reviews is a rough proxy for whether anyone is minding the shop, and it splits the field almost in half.
Both are easy things to check before you call, and both are filters we apply when building the city rankings on this site.
21 metros, ranked on rating weighted by review volume — the method this study argues for.
This is a sample of restoration firms that Google Maps surfaces for restoration search terms in 21 large US metros. It is not a random sample of the industry, and it under-represents rural markets and firms with no Google presence. Review text was sampled rather than exhaustive, and skews recent. Term-frequency analysis counts word occurrences and does not interpret sentiment or negation, so treat the direction of the findings as robust and the exact magnitudes as indicative. Google ratings and review counts are reproduced as attributed public data.
Local Restoration Directory (2026). "What 8,685 Restoration Reviews Reveal."
Analysis of 1,373 restoration firms across 21 US metros.
https://localrestorationdirectory.com/restoration-review-study/
Journalists and researchers are welcome to quote these findings with a link. If you'd like the aggregate figures broken out differently — by metro, by service type, or by rating band — get in touch and we'll share what we can.
Compare independently-ranked restoration companies in 21 metros — ranked on real Google reviews, never on payment.
Get our free homeowner claim checklist — what to photograph, when to file, what adjusters look for — plus occasional tips. Unsubscribe anytime.
Your checklist is ready right now — no inbox digging needed:
Open the claim checklist →