ChatGPT recommends water damage restoration companies it can check against third-party records, and for this trade the record it reaches for most is the Better Business Bureau. In the Cheers Market Baseline, ChatGPT cited bbb.org in 61.0% of its answers to "who are the best water damage restoration companies?" across 100 US and Canadian metros, against 14.5% of its answers across all 32 home-services questions we track. That is one question, asked in each metro once a week for the three weeks ending September 20, 2026, about 300 answers per engine, so read it as a strong signal for that question rather than a rule for every loss type. For a restoration brand with 20 or 200 markets, it means a stale or unclaimed BBB profile for any operating entity is a gap in the evidence ChatGPT actually reads.
Last verified: September 23, 2026.
The chart at the top shows the rest of ChatGPT's source mix on that question: restoration companies' own websites in 58.0% of answers (44.2% across all questions), Expertise.com in 54.3% (32.1%) and Angi or HomeAdvisor in only 8.3% (29.4%). Other directories and "best of" lists appeared in 45.3%, below their 59.9% rate across all questions. If you want to see which restoration companies each engine names in your own markets, the AI visibility check runs the same kind of question for one location in a few minutes.
Important
ChatGPT recommendations are outputs to measure, not rankings a restoration company can claim. Build verifiable market evidence, starting with the records ChatGPT cites on this question, then test whether it appears.
What the engines cite on the water damage question
The other engines split differently. Perplexity cited restoration companies' own websites in 97.9% of its answers, Expertise.com in 58.8%, the BBB in 48.8% and Thumbtack in 24.7%. Google AI Mode cited company websites in 95.3%, Google itself in 93.7%, Yelp in 24.3%, Thumbtack in 18.7% and Angi or HomeAdvisor in 16.7%. In the 153 answers where we captured Gemini's citations, it cited company websites in 90.8%, the BBB in 21.6% and Expertise.com in 19.0%.
The company's own site shows up on all four engines, and the BBB on three of them. The BBB is explicit about how its side works: "We publish profiles for Accredited and non-accredited businesses alike," and ratings are calculated "from a business's complaint history, how it responds to complaints, how transparent it is about its practices, and any licensing or government action against it." It also says "Accreditation dues do not buy a rating." For a restoration platform that has rolled up several local companies, every legacy legal entity may still have its own BBB profile, with its own complaint history, address and phone. Each one is a public record an engine can cite.
A lower review bar than other trades
The review bar on this question was among the lowest in the baseline. The restoration companies Google AI Mode named for water damage had a median of 80 Google reviews, against 226 across all 32 questions. In 42.5% of its answers at least one named company had fewer than 20 reviews, against 23.5% across all questions, and 6.4% of the companies it named showed under 4.0 stars, against 1.0% overall. In 18.7% of its water damage answers at least one named company sat under 4.0 stars, against 3.7% across all questions.
Cheers Market Baseline, Google AI Mode
A lower review bar
80
Median Google reviews of the restoration companies Google AI Mode named
Base: 1,107 rated appearances in 273 answers
226
Median across all 32 home-services questions
34,017 rated appearances
42.5%
Of answers named a company with under 20 reviews
23.5% across all 32 questions
6.4%
Of named companies showed under 4.0 stars
1.0% across all 32 questions
Cheers Market Baseline, August 31 to September 20, 2026: Google AI Mode place cards on the water damage restoration question and on all 32 questions, 100 US and Canadian metros. Fact IDs MB-T-WATER, MB-03, MB-04, MB-05.
| Finding | Value | Base |
|---|---|---|
| Median Google reviews of the restoration companies Google AI Mode named | 80 | 1,107 rated appearances in 273 answers |
| Median across all 32 home-services questions | 226 | 34,017 rated appearances |
| Of answers named a company with under 20 reviews | 42.5% | 23.5% across all 32 questions |
| Of named companies showed under 4.0 stars | 6.4% | 1.0% across all 32 questions |
These describe the companies that were named, not every restoration company in each market, so they show who got in rather than a cutoff. ChatGPT's map cards were not captured on this question in our window, so there is no ChatGPT rating figure to report, and Gemini's sample was too thin to publish. The practical read is still clear: on this question, a branch with a few dozen honest reviews sits in the same range as many of the companies that got named. Earning them steadily, by asking every eligible customer rather than only the happy ones, closes that gap faster here than it would in HVAC, where the median named business had 449.
National brands take a large share of this answer
National franchise brands appeared in 69.4% of the water damage answers that named a company, the highest share of any trade in the baseline (our pest control question came next, at 60.1%), and they were named first in 34.1%. They still filled only 28.6% of the named slots; independents and regional companies filled the rest. By engine, 74.7% of ChatGPT's answers included a national brand, 75.6% of Gemini's, 69.8% of Google AI Mode's and 55.8% of Perplexity's.
We count brands as a category and never name one, and the share leans high, because some answers mention national chains only to warn against them and franchise names are easier for our extraction to catch. For an independent restoration company, the number means the answer usually includes a national name and still has room for local ones. For a franchise location, brand presence gets the logo into the answer; the location still has to be the one named, which is the problem how franchise brands get recommended by ChatGPT works through.
Location pages get cited more than homepages
Water damage is the one question in the baseline where the homepage was not the most-cited page on business websites. Of 2,547 citations to restoration company websites, location or city pages took 33.4%, service-plus-city pages 26.4%, homepages 24.9% and plain service pages 7.2%. Across all 32 questions, homepages took 43.1%.
Cheers Market Baseline, water damage
Location pages beat the homepage
43.1%
Homepage share, all 32 questions
Share of citations to restoration company websites, by page type
Cheers Market Baseline, August 31 to September 20, 2026: 2,547 citation URLs on known restoration business domains, water damage restoration question, all four engines. Fact IDs MB-T-WATER, MB-20.
| Page type | Share of citations |
|---|---|
| Location or city page | 33.4% |
| Service-plus-city page | 26.4% |
| Homepage | 24.9% |
| Service page | 7.2% |
That puts the location page at the center of restoration visibility. It should help someone facing an active loss: a clear service area, a phone path that reaches the responding team, truthful availability, the loss types the branch handles, expected first steps, credentials and proof. The location-page checklist covers the broader multi-location standard.
Availability language is universal on this question. Mentions of 24/7 or after-hours service appeared in 100% of ChatGPT's water damage answers, 96.7% of Google AI Mode's, 99.3% of Gemini's and 85.6% of Perplexity's, while prices almost never came up. Google AI Mode linked an online booking page in 6.3% of its water damage answers, against 31.2% across all questions. The answer points people to a phone rather than a scheduler, so the location page's phone path does the job a booking button does in other trades.
Avoid copying the same water damage paragraph across 80 city pages. A useful Phoenix page might explain monsoon-related water intrusion, commercial drying capacity, local response boundaries and the team that takes the call. A Columbus page may have different storm, freezing or property-management proof. If there is no meaningful local difference and no accountable operating owner, the page is probably a doorway page rather than evidence.
Restoration market pages
Cloned page or loss proof
Cloned across 80 cities
Generic restoration page
- One water-damage paragraph, city name swapped
- 24-hour claim with no dispatch behind it
- Whole-state coverage, no boundary
- Credentials attributed to the parent brand
One accountable market
Loss-type market page
- Water, fire, mold, storm scoped separately
- Response boundary the branch can honor
- Up to 20 specific service areas, per Google
- Licenses named against the operating entity
Article body; service-area rules from Google Business Profile Help, Manage your service areas (support.google.com/business/answer/9157481), checked September 2026.
Service-area businesses also need a consistent coverage model, and Google's rules are short: "If you don't serve customers at your business address, remove your address from your Business Profile." Google allows up to 20 service areas and says the overall area "shouldn't be more than about 2 hours of driving time from where your business is based." None of that is a ChatGPT ranking factor. It is a durable standard for publishing honest local facts, and a useful ceiling for a restoration brand tempted to claim a whole state. The service-area coverage guide shows how to apply it.
Next step
Is AI recommending your business?
Find out how visible you are across ChatGPT, Gemini, Perplexity, and AI Overviews.
Separate proof by loss type
"Restoration" is too broad for many customer questions. A company may be excellent at residential water mitigation and lack the equipment, certification or staff for large commercial fire losses, and the public evidence should keep those differences visible.
For water damage, publish the response boundary, extraction and drying scope, equipment and monitoring process, documentation expectations and market-level job proof. For mold work, explain containment, assessment boundaries and when a separate environmental professional is involved. The EPA's advice to anyone hiring a mold contractor is to "make sure they have experience cleaning up mold, check their references". That is not an AI ranking signal. It is a good editorial test for whether the market page helps a cautious customer verify who will do the work.
Let ChatGPT reach the evidence
OpenAI's crawler documentation names one agent for search. OAI-SearchBot "is used to surface websites in search results in ChatGPT's search features," and sites opted out of it "will not be shown in ChatGPT search answers." It can take about 24 hours after a robots.txt change for OpenAI's systems to adjust. Check robots.txt, noindex rules, canonical tags and firewall rules on each canonical market page; the AI crawler guide separates search discovery from training controls.
What to do this week
Pick five priority markets. For each, find every BBB profile tied to the operating entity or any business it absorbed, and check the name, address, phone, license details and the response to each open complaint. Then open the market's location page and confirm it names the loss types the branch handles, the response boundary and a phone that reaches the responding team. Finally, run "who are the best water damage restoration companies?" with the market's city on ChatGPT and Google AI Mode, note which companies and which sources come back, and assign each gap an owner:
- A BBB profile, directory listing or Google Business Profile with the wrong facts goes to whoever owns that market's listings.
- A location page with no loss types, boundary or credentials goes to marketing, with operations signing off on what the branch can honor.
- A market where the named companies have more reviews goes to the branch manager as a review-request habit, not a campaign.
Do not expand to the full portfolio until the first five markets survive a public readback.
Methodology
Every week, Cheers asks ChatGPT, Gemini, Perplexity and Google AI Mode the same 32 home-services questions in 100 metro areas across the United States and Canada, and records every business each engine names, with its star rating, review count and map position where the engine shows one, and every source it cites. This analysis covers the three weeks from August 31 to September 20, 2026: 34,742 usable answers (9,434 from ChatGPT, 9,435 from Google AI Mode, 9,364 from Gemini and 6,509 from Perplexity) across 60 US and 40 Canadian metros. We excluded empty or cut-off answers, most of them from a Perplexity collection outage. Each question was asked once per city per week, so week-to-week changes include normal answer-to-answer variation. Business names were extracted automatically; a hand check of 40 answers found the extraction caught about three in four of the businesses a reader would count, so counts are lower bounds. Star ratings and review counts are the figures each engine displayed; for ChatGPT and Gemini we captured them from September 18 and September 14 onward. Findings describe associations in this sample, not how any engine ranks businesses.
Restoration is covered by one question in the set: "who are the best water damage restoration companies?" It drew 300 answers from ChatGPT, 300 from Google AI Mode, 300 from Gemini and 291 from Perplexity. Every restoration figure on this page describes answers to that question, not fire, mold or storm work, and not the trade as a whole. Google AI Mode showed ratings for 1,107 named company appearances in 273 of its answers (34,017 across all 32 questions). ChatGPT's map cards were not captured on this question during its capture window, and Gemini's rating sample was too small to publish. Gemini citation shares use the answers where we captured a citation. National franchise brands are counted as a category from an internal list; no brand is named.
Sources
Every external link below was opened and checked on September 23, 2026.
- Cheers Research methodology. Cheers Market Baseline, August 31 to September 20, 2026, 34,742 answers, aggregate only. Source of every restoration, engine, source, review, brand and page-type figure on this page.
- Frequently asked questions about BBB. Better Business Bureau, undated. Source of the three quoted lines on profiles, ratings and dues.
- Overview of OpenAI crawlers. OpenAI documentation, undated. Source of the quoted OAI-SearchBot description, the opt-out consequence and the roughly 24-hour adjustment window.
- Manage your service areas for service-area and hybrid businesses. Google Business Profile Help, undated. Source of the quoted address rule, the 20-area limit and the two-hour driving guidance.
- Mold remediation in schools and commercial buildings, chapter 1. United States Environmental Protection Agency, last updated February 18, 2026. Source of the quoted contractor-selection advice.
Amadeus Peterson is the CTO & Co-Founder of Cheers, the done-for-you local visibility partner for established home-service companies and PE-backed portfolios.