2026 Tri-Cities AI Search Visibility Study
We ran 405 standardized search executions across ChatGPT, Perplexity and Google Search to see which Tri-Cities home service businesses AI tools named, and whether Google showed AI Overviews.
Highlights
Executive Summary
In May and June 2026 we tested how AI search tools answer local contractor searches in Johnson City, Kingsport and Bristol, Tennessee. We wrote 45 standardized queries covering HVAC, plumbing, electrical, roofing and tree service across the three cities and ran each three times on ChatGPT (with web search), Perplexity and Google Search: 270 executions on ChatGPT and Perplexity combined, plus 135 Google searches, for 405 in total.
We matched the businesses named in the AI answers against the 68 businesses in our companion 2026 Tri-Cities Home Services Digital Visibility Study, which audited their websites and Google Business Profiles (GBP).
39 of the 68 businesses (57%) were named at least once by ChatGPT or Perplexity, and 23 were named on both. The other 29 were not named in any response we collected. Google showed no AI Overviews in any of the 135 searches.
Businesses with more Google reviews were named more often. Among businesses with 100 or more reviews, 21 of 27 (78%) were named, compared with 2 of 9 (22%) with fewer than 20. Named businesses averaged 352 Google reviews; businesses not named averaged 73. Named businesses also had higher GBP scores, while website scores and schema showed little difference. These are associations in one sample, not proof of cause.
Key Findings
Finding 1: No AI Overviews appeared in the Google searches we tested
Across 135 Google searches in an incognito browser session, we saw zero AI Overviews. Results showed Local Services Ads, the map pack or organic listings instead. For these queries, traditional local search remained the route to visibility. Google AI Mode was not tested.
Finding 2: 57% of the businesses were named by at least one AI tool
39 of 68 businesses were named at least once by ChatGPT or Perplexity: 23 by both, 7 by ChatGPT only and 9 by Perplexity only. The remaining 29 were not named in any response. Going unnamed in this sample does not mean a business is invisible to AI tools everywhere.
Finding 3: Businesses with more reviews were named more often
Named means mentioned at least once in the ChatGPT or Perplexity responses we collected. These are business-level proportions, not the chance of appearing in any one answer.
| GOOGLE REVIEW COUNT | BUSINESSES NAMED | NAMED / TOTAL |
|---|---|---|
| 200+ | 77% | 10 / 13 |
| 100-199 | 79% | 11 / 14 |
| 50-99 | 58% | 7 / 12 |
| 20-49 | 50% | 8 / 16 |
| Fewer than 20 | 22% | 2 / 9 |
The table covers 64 businesses with usable review-count data, 38 of them named. Four businesses had no review count in our audit data; one of them, Rowan Tree Care, was named. The 100 to 199 band had the highest proportion (79%), and the rate fell steadily below 100 reviews, from 58% to 50% to 22%. This is an association, not a guaranteed review threshold.
Finding 4: Named businesses had higher GBP scores
Named businesses averaged 86.5 on our GBP Health Score, versus 77.0 for businesses not named. Website Health Scores were close: 65.9 versus 64.3. These differences do not show cause, and they do not tell us where AI tools get their information.
Finding 5: Schema showed no clear link to being named
Businesses with LocalBusiness schema were named 57% of the time (20 of 35), versus 58% without it (19 of 33). The two groups also differed in reviews, averaging 357 and 104, so this comparison cannot isolate schema's own effect.
Finding 6: Mention rates varied by industry
| INDUSTRY | NAMED / TOTAL | SHARE NAMED | AVG REVIEWS (NAMED) | AVG REVIEWS (NOT NAMED) |
|---|---|---|---|---|
| HVAC | 10 / 14 | 71% | 658 | 26 |
| Plumbing | 8 / 12 | 67% | 472 | 147 |
| Roofing | 8 / 15 | 53% | 117 | 30 |
| Electrical | 6 / 12 | 50% | 279 | 64 |
| Tree Service | 7 / 15 | 47% | 64 | 95 |
HVAC had the highest share of businesses named (71%) and tree service the lowest (47%). Tree service also ran against the overall pattern: named tree service businesses averaged fewer reviews (64) than those not named (95). Businesses in more than one trade are counted once, under the industry listed in the methodology.
Finding 7: Perplexity named local businesses more often
Across all responses, Perplexity named the audited businesses 188 times and ChatGPT 135 times. In our review, Perplexity more often cited the businesses' websites, profiles and directory listings, while ChatGPT's answers varied more between runs. We did not measure citation behavior numerically.
Businesses Named on Both Platforms
23 of the audited businesses were named by both ChatGPT and Perplexity. Counts combine every spelling of a business's name found in the responses. Under our scoring formula each of these businesses scores 55 out of 100, because the formula rewards appearing on a platform rather than how often, so the counts show differences the score does not.
| BUSINESS | INDUSTRY | CHATGPT | PERPLEXITY | TOTAL |
|---|---|---|---|---|
| Steve Huff Plumbing, Heating & Air | Plumbing / HVAC | 7 | 20 | 27 |
| Stateline Services | Multi-trade | 11 | 10 | 21 |
| Austin's Plumbing LLC | Plumbing | 8 | 11 | 19 |
| Rowan Tree Care | Tree Service | 4 | 12 | 16 |
| ATC Cooling & Heating | HVAC | 8 | 8 | 16 |
| Tennessee Plumbing | Plumbing | 6 | 8 | 14 |
| Johnny's Electrical & HVAC | Electrical | 2 | 11 | 13 |
| Quality Heat & Aire | HVAC | 7 | 6 | 13 |
| Hometown Plumbing, Electrical & HVAC | Multi-trade | 2 | 10 | 12 |
| Vanderpool Roofing LLC | Roofing | 6 | 6 | 12 |
| Pink Duck Plumbing | Plumbing | 7 | 5 | 12 |
| Tri-City Air Conditioning Co. | HVAC | 8 | 4 | 12 |
| Thompson Electrical Service | Electrical | 3 | 8 | 11 |
| TriCities Electric | Electrical | 2 | 8 | 10 |
| Precision Electrical Company | Electrical | 2 | 7 | 9 |
| McClain Roofing and Siding | Roofing | 3 | 5 | 8 |
| Wommacks Mechanical | HVAC | 5 | 3 | 8 |
| Palisade Roofing | Roofing | 7 | 1 | 8 |
| Smoky Peaks Plumbing | Plumbing | 4 | 3 | 7 |
| Mountain Air Inc. | HVAC | 5 | 2 | 7 |
| Bartlett Tree Experts - Johnson City | Tree Service | 4 | 2 | 6 |
| Redline Tree Service LLC | Tree Service | 1 | 4 | 5 |
| East Appalachian Electric | Electrical | 2 | 1 | 3 |
Sorted by total mentions. Counts are the totals recorded by our name extraction across all runs.
What Local Businesses Can Learn
1. Keep a steady process for earning honest reviews
Businesses in higher review-count bands were named more often in this sample. That makes reviews a useful part of a broader local visibility plan, not a guaranteed threshold or a replacement for fixing technical problems. Ask every customer for an honest review, without filtering by expected sentiment or offering incentives, and respond to feedback professionally.
2. Keep your Google Business Profile accurate and useful
Keep hours, contact details, website link, services and primary category current. Named businesses had higher GBP scores, which is a reason to review profile quality, but it does not prove profiles cause AI mentions. Choose the category that most accurately describes your main business.
3. Keep doing the local SEO fundamentals
For the Google searches we tested, visibility came from regular search results, not AI Overviews. Useful website content, accurate business information, crawlability and local relevance all still matter. AI Overviews and AI Mode are separate Google features, and our results do not predict how either will change, so keep checking real searches over time.
Methodology
Study Design: The 2026 Tri-Cities AI Search Visibility Study queried three AI platforms with standardized local home service search queries. The study covered five industries across three Tri-Cities markets: Johnson City TN, Kingsport TN, and Bristol TN.
Query Design: Three query templates were used for each industry-city combination: a "best of" query, a direct recommendation query, and a top-rated query. This produced 45 unique queries (3 templates × 5 industries × 3 cities). Each query was run three times per platform to measure response consistency. That totals 270 executions on ChatGPT and Perplexity and 135 on Google.
Platforms: ChatGPT (GPT-4o with web search enabled), Perplexity (default model with live web search), and Google AI Overviews (incognito Chrome, Johnson City TN geolocation). All queries were run within a two-week window in May–June 2026. Google AI Mode was not evaluated.
Cross-Reference Dataset: We extracted business names from each response with automated pattern matching followed by manual review, then matched them to the 68 businesses in the 2026 Tri-Cities Home Services Digital Visibility Study. A name was matched when it clearly referred to the same business, including spelling and suffix variants such as "Tennessee Plumbing Inc." and "Tennessee Plumbing Inc". Sentence fragments, street names, directory labels and businesses outside the audit were not matched. Businesses in more than one trade are counted once: Stateline Services, Steve Huff Plumbing, Heating & Air, A Breath of Fresh Air and Appalachian Mechanical Services under HVAC, and Hometown Plumbing, Electrical & HVAC and IDM Heating, Cooling, Plumbing under plumbing.
Scoring: AI Visibility Score (0–100): ChatGPT mention (20 pts), Perplexity mention (20 pts), Google AI Overview mention (30 pts), cross-platform bonus for two or more platforms (15 pts). Maximum attainable score given the Google AI Overview finding was 55.
Corrections in this edition
- Name variants are now matched consistently, adding five businesses named by at least one platform and raising the number named on both from 20 to 23.
- Businesses named: 39 of 68 (57%), not 34 (50%).
- Review-count, GBP, schema and industry figures were recalculated. Tree service, not roofing, now has the lowest share named.
- The both-platform table now lists all 23 businesses with counts combined across name variants.
- Findings are described as associations rather than predictors or causes.
Limitations
Point-in-time snapshot: AI responses change as models update. Results may not reflect current AI behavior.
Response variance: AI platforms do not return deterministic results. Three runs per query captured some variance but not the full distribution of possible responses.
Correlation only: Relationships between review count, GBP health, and AI visibility are correlational. The study does not establish causation.
Name extraction and matching: Automated extraction can miss or miscount names, and matching name variants involves judgment. Some Bristol responses described Bristol, England; those businesses were outside the audit and not matched.
Market scope: The study covers the Tri-Cities market only. Findings should not be generalized to other markets without independent research.
Casey Carmical
Owner, 1-FIND SERVICES – Johnson City, Tennessee
1-FIND SERVICES is a digital marketing agency based in Johnson City, Tennessee, serving the Tri-Cities region. This study was produced independently; no audited company paid to be included or excluded.