Across 174 Oklahoma home service websites audited on November 4 and 5, 2026, the average score was 48.3 out of 100, the median was 49, and the range ran from 23 to 82. The sites did best on technical health (73 on average) and worst on conversion (31) and trust signals (34). Roofing scored highest of the six trades at 52, and no trade averaged above 52.
What we did, and what this is not
We collected 206 websites of local home service companies from public business directory listings in 25 Oklahoma cities, and ran Realo's audit tool on each one on November 4 and 5, 2026. The tool could read pages on 174 of them. The other 32 block automated visitors or build their pages only with scripts, so they were left out of every number on this page. The tool crawled each site, loaded pages in a real browser (170 of the 174), and scored the site out of 100 across five areas. Some checks are measured directly, such as robots.txt, schema and page speed. Others are estimated from page wording.
This is our own first-party data. It is a directory sample, not a random draw of every contractor in Oklahoma, and it leans toward sites that are easy for a tool to read. It is not a ranking study and it doesn't measure traffic, calls or revenue. We aren't naming any business, because a score is a snapshot of a website and says nothing final about the company. We also don't set these results against our earlier Tulsa study of 20 sites, because the tool changed between the two runs.
The scores by area
Each area is scored out of 100, then averaged across the 174 sites.
The scores by trade
Roofing led, but the spread between trades was small. The spread inside each trade was far larger, from a low of 23 to a high of 82.
| Trade | Sites scored | Average score |
|---|---|---|
| Roofing | 31 | 52.0 |
| HVAC | 29 | 48.7 |
| Plumbing | 28 | 47.7 |
| Electrical | 30 | 47.4 |
| Pest control | 27 | 47.1 |
| Remodeling | 29 | 46.5 |
What most of them did well
The basics were mostly in place. Out of the sites the tool could test on each point:
- 149 of 174 let AI crawlers in. The rest blocked them through robots.txt or a firewall.
- 140 of 163 had no near-duplicate town pages of the kind Google treats as doorway pages.
- 136 of 174 kept their layout stable while loading.
- 117 of 139 stayed on topic, with content that matched the service they sell.
That is the work most web designers get right, and it explains why technical health was the highest score.
What almost none of them had
The gaps were the same across all six trades. These are the checks where the largest number of sites scored zero:
| What the audit looked for | Sites that scored zero |
|---|---|
| A written case study with the problem, the fix and the result | 161 of 171 |
| A calculator, estimator or diagnostic tool | 156 of 169 |
| Services with prices in structured data | 144 of 169 |
| Service areas marked up in structured data with real places | 142 of 171 |
| A named person as author of guides and articles | 126 of 173 |
| A comparison table on any page | 117 of 169 |
| Prices or cost ranges shown on the site | 88 of 164 |
| Proof such as license or reviews on the first phone screen | 82 of 170 |
Most of these fixes are small. Marking up your service area, listing a price range and adding a named author take an afternoon each. An estimator or a real job write-up takes longer, and almost no one in the group had either.
Why the conversion score was lowest
Conversion averaged 31, the lowest of the five areas. The sites mostly said what the company does and where. Few put proof in front of a visitor on a phone, and the tool found no estimator or calculator on 156 of 169 sites. A homeowner who can't tell whether you're licensed, reviewed or in their price range goes back to the results page.
Speed and the call button were weak spots too. Only 30 of 173 sites loaded their main content in 2.5 seconds or less on a simulated mobile connection, which is Google's bar for a good score. Only 27 of 170 had a tap-to-call button and a clear main action visible on a phone without scrolling. Both matter when someone is trying to reach a roofer or plumber with one thumb.
What this means for AI search
AI tools name businesses whose facts they can read and check. Structured data that names your towns and prices, a named author, a visible license and a comparison table are all facts a tool can quote. Most of the 174 sites offered none of them, which leaves room for any contractor who adds them. Our guide on getting recommended by ChatGPT covers the full list, and the AI search page explains how we work on it.
How to check your own site
Start with the table above. If your site has none of those items, you have the same gaps as most of the group. Realo's free audit runs the same checks on your site and returns scores plus a ranked list of fixes, reviewed by a person.
Common questions
How was the Oklahoma contractor website audit done?
Realo ran its own audit tool on 206 Oklahoma home service company websites on November 4 and 5, 2026. The tool could read pages on 174 of them, and those 174 are scored here. Each site was crawled and scored out of 100 across five areas. The scores are Realo's own measurements, not Google's.
Which businesses were audited?
The sites came from public business directory listings in 25 Oklahoma cities across six trades: roofing, HVAC, plumbing, electrical, pest control and remodeling. It is a directory sample, not a random draw of every contractor in the state. Realo isn't naming the businesses, because a score is a snapshot of a website, not a verdict on the company.
Why were 32 sites left out?
The audit could read no pages on 32 sites, usually because the site blocks automated visitors or builds its pages only with scripts. Leaving them out keeps the scores honest, but it also means the sample leans toward sites that are easy to read.
Which trade scored highest?
Roofing averaged 52 out of 100. HVAC averaged 48.7, plumbing 47.7, electrical 47.4, pest control 47.1 and remodeling 46.5. The gap between trades is small next to the gap between the best and worst sites, which ran from 23 to 82.
Is this a ranking or traffic study?
No. The audit measures what is on a website and how it is built. It doesn't measure Google rankings, traffic, calls or revenue, and a high score doesn't promise more leads.
Why doesn't this compare with the earlier Tulsa study?
The audit tool changed between the two runs, so the averages aren't measured the same way. The Tulsa study covered 20 sites and averaged 54. We treat the two as separate snapshots rather than a trend.
Can Realo check my site the same way?
Yes. The free audit runs the same checks on your site and returns scores plus a ranked list of fixes, reviewed by a person.
Want to see where your site lands?
Send us your website address and we'll run the same audit on it, free, with no obligation.
Get my free audit