Original Research
We audited 209 local business websites. The basics are fine. Everything else is not.
Plumbers, HVAC companies, roofers, and med spas across 12 US metros, pulled from live Google results and tested one by one. The surprising part is what is not broken. The useful part is what is.
Led by AJ Rodriguez Sanchez, Founder and Lead StrategistPublished
209
websites tested
12
US metros
4
local trades
10
checks per site
Three numbers that should not exist in 2026
21%
failed to load at all
44 of 209 sites never served a homepage to a first-time visitor: bot protection gone wrong, dead servers, DNS failures. Every one of those visits was a customer who went elsewhere.
41%
slower than 2 seconds
More than four in ten loadable sites took over 2 seconds just to start responding. 13% took over 3. The median local business site responds in 1.64 seconds, twice what Google considers good.
23%
have no analytics at all
Nearly a quarter of local businesses are flying blind: no tag manager, no analytics, no ad pixels. They could not tell you which page earns their calls if you asked.
The interactive map
Signal Terrain: every light is one of the 209
Each signal is one anonymized site from the study, clustered at its metro. Cream passed every check, dune means one or two failures, blaze means three or more, or a site that never loaded at all. Hover a signal to see exactly what failed, or color the map by speed, contact path, or measurement.
Anonymized records only: no business names, no addresses, metro-level positions. The full dataset behind the map is the one published in the methodology below.
The surprise
The textbook SEO checklist is not the problem anymore
If you picture the average local business website as a tagless mess, update the picture. Title tags, meta descriptions, canonicals: nearly universal. Twenty years of checklists and plugins worked. The failures moved somewhere else.
They moved to the things a checklist cannot fix with a plugin: server speed, measurement, and giving a ready customer an obvious way to reach you.
Missing a meta description
2%
Missing a canonical tag
4%
Missing structured data (schema)
7%
Share of loadable sites failing each basic check. Lower is better, and lower is what we found.
Where the rot actually is
Ranked by frequency across the 165 sites that loaded. Read the top two together: four in ten sites are slow, and four in ten give no easy way to make contact. Both cost calls on the same visit.
No contact form on the homepage
41%
The single most common failure. The phone number is there; the easy path is not.
Slower than 2 seconds to respond
41%
Measured from a real connection, not a lab.
15+ images, none lazy-loaded
28%
Heavy pages that punish mobile visitors on cellular connections.
No analytics detected
23%
No measurement means no way to know what marketing works.
No tap-to-call on the homepage
16%
A phone number you cannot tap, on devices people hold in their hands.
Visibly unmaintained (stale footer)
8%
Copyright years stuck in 2023 or earlier. Customers notice, fair or not.
Every trade has its own failure mode
AC repair
n=47Slowest trade in the study: over half of HVAC sites take more than 2 seconds to respond, on keywords where a click costs $20 to $80.
Med spa
n=46A visual trade with a conversion problem: a third of med spa homepages offer no tap-to-call and no form.
Plumber
n=31Most likely to look abandoned: 16% of plumber sites show a copyright year of 2023 or older.
Roof repair
n=41The measurement blackout trade: more than a third of roofing sites run no analytics whatsoever.
Page one is not clean either, but page two is worse
Sites ranking on page one fail less often than sites buried on page two and beyond. Correlation is not causation, but the direction of every gap is the same, and none of these are ranking factors a plugin installs.
Page 1 (positions 1-20)
n=99Page 2+ (positions 21-50)
n=66If you own one of these sites
The fixes that pay back first
None of the top failures require a rebrand or a rebuild to start. They require an afternoon of honest testing against the standards we audit to, and a willingness to treat the website as the front desk it actually is.
See what the full audit covers- 1
Answer the phone-path question: can a phone user reach you in one tap, and can anyone reach you without calling at all?
- 2
Time your own site on a real phone on cellular. Over two seconds to first response is the top of the market failing, not a technicality.
- 3
Install measurement before spending another marketing dollar. 23% of your competitors cannot tell what works; that is an advantage if you can.
- 4
Check the date in your own footer. If it says 2023, everything else you claim about your business gets discounted.
Methodology
How this study was run
Sample: live Google results for plumber, AC repair, roof repair, and med spa in Phoenix, Nashville, Tampa, Kansas City, Indianapolis, Las Vegas, Portland, Charlotte, Denver, Atlanta, Dallas, and Houston, positions 1 through 50, pulled November 6-7, 2026. Directories, aggregators, national franchises, lead-generation networks, and out-of-market results were removed by hand. 209 local business homepages remained and were each tested for ten signals: response time, page weight, title and meta hygiene, canonical tags, viewport, structured data, Open Graph tags, analytics presence, contact form, tap-to-call, image lazy-loading, and footer freshness.
Honesty notes: the 44 sites counted as failed to load include 23 that returned 403 to a first-time automated visitor, 16 that timed out or failed DNS, and 5 that returned server errors. Some 403s are deliberate bot protection, and we count them anyway, for the reason explained in the FAQ. Percentages for load failures use the full 209-site sample; all other percentages use the 165 sites that loaded. No site is named in this study. The same checks power our published audit methodology.
Questions about the study
How were the 209 websites chosen?
We pulled live Google results for plumber, AC repair, roof repair, and med spa across 12 US metros (positions 1 through 50), removed directories, franchises, and lead-generation networks, and audited every remaining local business homepage: 209 sites, tested November 6-7, 2026.
Does a 403 error really count as a failure?
We think so, with a caveat. Some 403s are bot protection doing its job. But the same rules that block a research crawler frequently block AI crawlers and other first-time automated visitors, and none of the affected businesses can see the difference in their logs, because 23% of them run no analytics at all.
Why does page speed get so much weight?
Because these are emergency and appointment trades. A homeowner with a leaking pipe or a broken air conditioner in Phoenix in summer is not waiting four seconds. Speed is not an SEO checkbox for these businesses; it is the difference between the call and the back button.
Can you check my site against this dataset?
Yes. The free screen covers the same checks this study used. The full Growth Audit goes far deeper: crawl, render-gap testing, SERP and AI-citation research, and a prioritized 90-day roadmap.
Want to know which of these your site fails?
The Growth Audit runs every check in this study, plus the layers a homepage screen cannot reach: crawl, render gap, SERP reality, and AI citation testing. $1,500 flat, credited to month one if we continue.
Book the Audit