THE SHORT ANSWER
On a phone, about 1 in 10 Florida pool service websites shows no phone number and no call, quote or contact button on the first screen. That is 11 of 114 independent sites listed on Google Maps in four Florida metros (10%, 95% CI 5%–16%), each checked by hand on a phone-size screenshot taken Oct 30, 2026. Among 72 sites found through state registries, the share was 17%.
Key findings
11 of 114
11 of 114 sites (10%, 95% CI 5%–16%) show a phone visitor no phone number and no call, quote or contact button before they scroll. Only 35% put both a tap-to-call link and a quote or contact button on the first screen.
47 of 100
The typical site scores 47 out of 100 on the index. 24% land in the lowest band ("hard to buy from") and 2 of 114 reach 80.
82% vs. 24%
82% of sites say something about cost, but only 24% give a number. 43% mention insurance; 8% say what kind.
62% vs. 18%
Outside the index, 62% of the 29 top-scoring sites have a separate service-area page, against 18% of the rest.
17%
Sites found through Florida’s state registries look the same or worse: a median 43 on the four pillars without speed (Maps: 44), and 17% of their first screens are dead ends.
Method: a 100-point index, fixed before we saw the data
The index measures how easy a website makes it for a homeowner on a phone to reach, trust and choose a pool service business. It scores readiness from features we can observe; it does not measure actual sales or leads, which we cannot see.
A crawler opened each site’s homepage in a phone-sized browser (390×844 CSS pixels, touch enabled), recorded what was visible on the first screen before scrolling, and read up to six inner pages (contact, about, services, service area, FAQ, reviews) plus a blog index. "Visible" means at least 60% of the element’s width and half its height on the first screen, not hidden and not covered. Speed comes from Google Lighthouse lab tests on a simulated mid-range phone (median of three runs).
Weights and thresholds were written down on Oct 30, 2026, before any crawl results were seen, and frozen in the scoring script
(sha256 of the weights: 702b0f9b8934b3f7…). Each item is 0 or 1 unless noted; a pillar score is
the weighted share of its items, and the total is the weighted mean of the pillars.
| Pillar | Points | Items (points) |
|---|---|---|
| Reach me fast | 25 | Tap-to-call link on first screen 7 · Phone number visible on first screen 3 · Quote/contact/book button on first screen 5 · Call button that stays on screen 3 · Form on first screen 2 · Contact form anywhere 2 · Text, chat or online booking 2 · Clickable email link 1 |
| Trust | 20 | License number shown 3 · Says "licensed" 2 · Says "insured" 3 · Reviews or testimonials on site 4 · Street address 2 · Years in business 1 · Photos of real people / team 2 · Owner named 1 · Association or BBB badge 1 · Email on own domain 1 |
| Answers | 20 | Areas served 3 · Contract / how to cancel 3 · How to get started 3 · How fast you start / respond 3 · Equipment repair 2 · Green pool / one-time clean 3 · Visit reports / communication 3 |
| Price clarity | 15 | Any $ price for service 7 · What it costs 3 · What's included 3 · Chemicals included 2 |
| Mobile & speed | 20 | Mobile viewport set 2 · No sideways scrolling at phone width 2 · Body text 16 px+ 2 · Few too-small tap targets 2 · Lab LCP (scaled) 6 · Lab TBT (scaled) 3 · Lab CLS (scaled) 3 |
Questions score yes = 1, partial = 0.5. Lab speed: LCP full credit at ≤ 2.5 s and none at ≥ 6 s; TBT (a lab stand-in for responsiveness) ≤ 200 ms to ≥ 600 ms; CLS ≤ 0.10 to ≥ 0.25, linear in between. A site needs 80% of the total weight observed to be scored; a pillar needs 60%. Sites without Lighthouse data are scored on the other four pillars and flagged.
Evidence behind the items: phone calls from mobile search [1, 2], speed of follow-up [3], reviews [4, 5], up-front disclosure [6], cost transparency [8–10], page speed [11–14] and mobile share of traffic [15]. Florida law requires a contractor’s license number in every offer of services and advertisement [7]; cleaning that does not affect the pool’s structure or equipment needs no state license [7], so that item carries 3 of 100 points.
Sample and coverage
Primary frame: what a homeowner finds on Google Maps. 251 independent pool service businesses from Google Maps searches in four Florida metros (collected Oct 27, 2026; national chains, franchises and directories removed; one row per domain). 167 have their own website. We crawled all of them; 114 loaded and were measured. The other 53 were excluded rather than scored zero: 19 HTTP 403, 10 robots unavailable, 6 dns fail, 5 HTTP 404, 4 bot blocked, 2 robots disallow, 2 HTTP 5xx, 1 conn reset, 1 empty page, 1 timeout, 1 site error, 1 parked or down. Most of these are firewalls that turn away automated visitors; we did not work around them.
Comparison frame: found through public registries. A stratified sample of Florida Division of Corporations entities with a pool word in their name plus all active state pool/spa contractor licensees, with websites found by matching names to domains (no search engine or map data). 82 sites, 72 measured. Only a handful overlap with the Maps frame. This frame shows whether the picture depends on how the businesses were found. We never pool the two frames.
| Google Maps frame | Registry frame | |
|---|---|---|
| Own websites in frame | 167 | 82 |
| Measured (scored) | 114 (68%) | 72 (88%) |
| With Lighthouse speed data | 110 | 70 |
| Lighthouse failed / not yet run | 4 / 0 | 2 / 0 |
| By metro | Tampa Bay 72, Orlando 20, Fort Lauderdale 15, Fort Myers 7 | Southwest FL 19, South Florida 17, Rest of Florida 14, Tampa Bay 12, Orlando 10 |
| By type (Maps category / license class) | Pool cleaning service 71, Pool contractor 28, Pool repair service 15 | registry 48, CPC-B 10, CPC-A 5, CPC-C 8, RP-C 1 |
The Maps frame leans toward Tampa Bay (72 of 114) because one county was searched more finely; metro results are shown only where n ≥ 10. "Registry" in the type row means a business entity without a pool license.
1 in 10 first screens gives a caller nothing to tap
A homeowner who searched for pool service on a phone and tapped through to a site sees one screen first. On 11 of 114 sites that screen shows no phone number and no call, quote or contact button. 8 of these sites do have a tap-to-call link somewhere else; it just isn’t where the visitor is looking.
A smaller leak: 10 of 114 sites show the number on the first screen as plain text that can’t be tapped, so the visitor has to copy it by hand.
How we counted. The crawler flags a first screen with no phone number, no call, quote, contact or booking button and no form: 20 sites (18%, CI 12%–26%). We then checked each of them by hand on the screenshot, and on the page code where a layer hid part of the screen. 6 had a way to act after all (a button under a pop-up or loading layer, an unlabeled floating button, a number in a desktop-only header), and 3 could not be judged because a banner or an unfinished page hid the screen. That leaves 11; counting the unresolved ones too gives 14 (12%). Among sites without a pop-up or banner covering a fifth of the screen, the hand-checked share is 10% (CI 5%–17%).
Share of 114 sites (Google Maps frame), first screen at 390×844 before scrolling. 95% CI for the hand-checked share: 5%–16%.
Source: Ruberio crawl, Oct 30, 2026; first-screen fields agreed with a manual check on 97–100% of a 29-site holdout; every flagged dead end checked by hand on its screenshot.
Illustration re-created from measured shares, not a real site. Solid = at least half of the group has it on the first screen; dashed = fewer. Percentages: share of sites in each group.
Source: Ruberio crawl, Google Maps frame; groups by index score.
The typical site scores 47 out of 100, and price is the weakest pillar
Number of sites per 10-point score range, Google Maps frame. Vertical line = median (47).
Source: Ruberio Conversion Index, Oct 30, 2026.
| Band | Sites | Share (95% CI) |
|---|---|---|
| 80-100 · ready to convert | 2 | 2% (0%–6%) |
| 60-79 · solid, with gaps | 28 | 25% (18%–33%) |
| 40-59 · makes the customer work | 57 | 50% (41%–59%) |
| 0-39 · hard to buy from | 27 | 24% (17%–32%) |
Price clarity is the lowest pillar because most sites show no dollar amount for service (24% do). Mobile basics are mostly fine (89% set a mobile viewport, 97% avoid sideways scrolling); the mobile pillar loses its points on lab speed, where 4 of 110 sites with speed data show their main content within 2.5 seconds.
Dot = median pillar score, bar = middle half of sites (25th–75th percentile), out of 100. Mobile & speed only for sites with Lighthouse data.
Source: Ruberio Conversion Index, Google Maps frame.
Median scores out of 100; only groups with n ≥ 10. Type = the business’s Google Maps category.
Source: Ruberio Conversion Index, Google Maps frame.
Pool contractors (builders and remodelers) score a median 60 on trust and 10 on price clarity; cleaning services score 35 and 37. Florida law requires licensed contractors to put their license number in every offer of services and advertisement [7]; 64% of the contractor sites in this frame show one. Of all sites that say "licensed", 42% don’t show the number.
Most sites raise the question and leave out the answer
Share of 114 sites (Google Maps frame). "Mentioned": e.g. an estimate at no charge or "call for pricing" for cost; "insured" without saying what kind; the word "contract" without terms.
Source: Ruberio crawl, rule-based reading of homepage plus up to six inner pages. Indicative: only "yes" answers were checked by hand (15 sites); areas, cost, license and repair are the most reliable.
| Question | Mentioned | Answered | Gap |
|---|---|---|---|
| What it costs | 82% | 24% | 59 pts |
| How to get started | 69% | 22% | 47 pts |
| Chemicals included | 66% | 23% | 43 pts |
| Insured | 43% | 8% | 35 pts |
| How fast you start / respond | 66% | 32% | 33 pts |
The median site answers 4 of 12 questions with a clear yes (middle half: 2–6). The gap is widest on what it costs: 82% of sites touch the topic, 24% answer it.
The top quarter answers more of everything; the difference is largest on how fast you start / respond (72% vs 19%) and visit reports / communication (62% vs 16%). The questions are part of the index, so this shows where the top quarter earns its points, not why.
How we framed the questions: we checked every site against 12 questions a homeowner needs answered before calling a pool service company. The list and the grouping are our framing of the decision (can you serve me, what will it cost, can I trust you, how does it work); we did not survey homeowners. Consumer guidance points the same way: Palm Beach County’s consumer affairs office tells homeowners to confirm a pool contractor’s certification and to get at least three written quotes [16].
What the top quarter does differently
We split the 114 sites at the 75th percentile of the index (score ≥ 62, 29 sites) and compared how often each feature appears. Items that are part of the index are more common in the top quarter by construction; for those, the ratio only tells which items separate sites most. Features outside the index are the interesting part, but with 29 sites in the top group the intervals are wide and none of this shows cause and effect.
| Index item | Top | Rest | Ratio (95% CI) |
|---|---|---|---|
| Form on first screen | 14% | 2% | 5.9× (1.1–30.3) |
| Call button that stays on screen | 72% | 18% | 4.1× (2.5–6.8) |
| Lab LCP (scaled) | 7% | 2% | 2.8× (0.4–18.9) |
| Any $ price for service | 45% | 16% | 2.7× (1.5–5.1) |
| Text, chat or online booking | 41% | 18% | 2.3× (1.2–4.4) |
| Says "insured" | 72% | 33% | 2.2× (1.5–3.2) |
| Tap-to-call link on first screen | 90% | 41% | 2.2× (1.6–2.9) |
| Association or BBB badge | 62% | 32% | 2.0× (1.3–3.0) |
| Outside the index | Top | Rest | Ratio (95% CI) |
|---|---|---|---|
| Specific insurance (liability, workers comp) | 17% | 5% | 3.7× (1.1–12.7) |
| Separate service-area page | 62% | 18% | 3.5× (2.0–6.0) |
| First-screen button styled as a button | 86% | 42% | 2.0× (1.5–2.7) |
| Custom or unidentified platform | 48% | 25% | 2.0× (1.2–3.3) |
| Separate reviews page | 45% | 24% | 1.9× (1.1–3.3) |
| Links to a blog | 55% | 29% | 1.9× (1.2–3.0) |
| Says "family-owned" | 62% | 34% | 1.8× (1.2–2.7) |
| First-screen button promises no charge | 45% | 25% | 1.8× (1.0–3.1) |
| LocalBusiness schema markup | 76% | 45% | 1.7× (1.2–2.3) |
| Feature (outside the index) | Top quarter | Rest | Ratio (95% CI) |
|---|---|---|---|
| Built on WordPress | 21% | 42% | 0.5× (0.2–1.0) |
| Built on a site builder (Wix, GoDaddy, Duda ...) | 31% | 33% | 0.9× (0.5–1.8) |
| Any analytics tag | 62% | 48% | 1.3× (0.9–1.8) |
| Google Ads tag | 28% | 25% | 1.1× (0.6–2.2) |
| Links to 2+ social profiles | 72% | 60% | 1.2× (0.9–1.6) |
| Pop-up or banner covers 20%+ of first screen | 17% | 18% | 1.0× (0.4–2.5) |
| Chat widget | 17% | 13% | 1.3× (0.5–3.5) |
Top sites are built to be found by place: a separate service-area page (62% vs 18%) and LocalBusiness schema markup, structured data that tells search engines the business name, area and hours (76% vs 45%), are both clearly more common. They also keep calling within reach: a call button that stays on screen shows one of the clearest gaps among index items. Tools are not what separates them: chat widgets, analytics tags, links to social profiles and Google Ads tags appear at similar rates in both groups. WordPress sites are less common in the top quarter (21% vs 42%, though the interval reaches 1). Median index score by platform: WordPress 44 (n = 42), site builders 47 (n = 37), custom or unidentified 51 (n = 35).
Businesses found through public registries look the same or worse
Google Maps shows businesses that keep a Business Profile, which may favor firms that work on their marketing. To check, we ran the same crawl and index on 72 sites found through Florida’s corporation and contractor-license records. The registry sites are spread over more of Florida (including South Florida and the Southwest) and are more often licensed contractors.
| Measure | Google Maps frame (n = 114) | Registry frame (n = 72) |
|---|---|---|
| Median score, four pillars without speed | 44 | 43 |
| Middle half, four pillars | 35–59 | 23–56 |
| Below 40 on four pillars | 41% (33%–50%) | 46% (35%–57%) |
| Reach me fast (median pillar score) | 48 | 38 |
| Trust (median pillar score) | 45 | 40 |
| Answers (median pillar score) | 48 | 48 |
| Price clarity (median pillar score) | 33 | 27 |
| Dead-end first screen | 10% (5%–16%) | 17% (10%–27%) |
| Tap-to-call on first screen | 54% (44%–62%) | 47% (36%–59%) |
| Quote/contact button on first screen | 57% (48%–66%) | 57% (45%–68%) |
| Any $ price for service | 24% (17%–32%) | 10% (5%–19%) |
| License number shown | 33% (25%–42%) | 31% (21%–42%) |
| Reviews or testimonials on site | 59% (50%–67%) | 39% (28%–50%) |
| Questions answered clearly (median of 12) | 4 | 3 |
Registry-found sites score the same or worse on every measure in the table: the first-screen gap is not a quirk of businesses that keep a Google profile. Scores here leave out Mobile & speed for both frames, because the registry sites were speed-tested later and not all of them could be measured; the four-pillar score uses the pre-registered weights of the other pillars.
The ranking holds when the weights change
The weights are partly judgment, so we tested whether the conclusions depend on them, exactly as pre-registered: equal pillar weights, leaving out each pillar in turn, and 2,000 random draws where every pillar weight (and, separately, every item weight) is multiplied by a factor between 0.5 and 1.5. The pre-set rule: a finding is robust if the median rank correlation with the main ranking is at least 0.85 and it holds in the equal-weight and every leave-one-out variant.
| Variant | Rank corr. (ρ) | Top-quarter overlap |
|---|---|---|
| Equal weights | 0.98 | 97% |
| Without Reach me fast | 0.86 | 79% |
| Without Trust | 0.91 | 79% |
| Without Answers | 0.96 | 93% |
| Without Price clarity | 0.94 | 86% |
| Without Mobile & speed | 0.97 | 97% |
| Monte Carlo, 2,000 draws | As coded | Corrected |
|---|---|---|
| Pillar weights: median ρ | 0.97 | 0.98 |
| Pillar weights: 5th pct. ρ | 0.95 | 0.95 |
| Item weights: median ρ | 0.99 | 0.99 |
| Median 90% rank range (places) | 17 | 17 |
| Top-quarter sites staying top in 90% of draws | 79% | 79% |
| Sites keeping their band in 90% of draws | 70% | 70% |
The headline findings do not depend on the weights at all: the dead-end share is a direct count, checked by hand. The median score stays in the 40–59 band under equal weights, every leave-one-out variant and 100% of random draws (range of the median across draws: 46–50); the share below 60 stays between 73% and 76%. Individual sites move more: 30% of sites sit close enough to a band edge that their band changes in more than one draw in ten. Band labels for single sites are therefore approximate. Mapping the "quote/contact/book button" item to any first-screen button instead (including call buttons) gives a ranking with ρ = 0.99.
Limitations
- Readiness is not conversion. We measured what a site offers a visitor, not how many visitors call. The weights rest on studies of retail, B2B, HVAC and roofing customers, not pool service.
- Automated reading. First-screen call and button fields and the license number agreed with a manual check on 97–100% of a holdout sample. That check could not tell a service price from other dollar amounts: on the full set, 7 of the 34 sites the crawler flagged for a price showed only financing, do-it-yourself or construction amounts or a repair-approval limit. We therefore use the hand classification from our pricing study (No. 04): 24% instead of 30%, registry 10% instead of 15%. Every first screen the crawler flagged as a dead end was checked by hand; 6 of 20 had a button the crawler could not see. The 12 questions are indicative: only "yes" answers were spot-checked (15 sites), and missed answers were not measured. Text inside images (a phone number in a logo) is invisible to the crawler.
- Lab speed from the Netherlands. Lighthouse simulates a mid-range phone on a slow connection from a lab; it is not what real Florida visitors experience, and absolute LCP values are high (median 8.1 s in this frame).
- Frame. The Maps frame comes from one set of searches, leans toward Tampa Bay and covers four metros; 53 sites (32%) could not be measured, mostly because firewalls blocked automated visits. Blocked sites may differ from measured ones (often larger, managed hosting).
- One moment in time. Sites were crawled on Oct 30, 2026. Pop-ups, seasonal banners and A/B tests can change a first screen from day to day.
- What would change the conclusion: a crawl that renders first screens differently (other phone sizes, a dismissed cookie banner) or a larger, statewide frame. The hand-checked dead-end share without sites that showed a pop-up is 10%; counting the 3 screens we could not judge as dead ends gives 12%.
What an owner can do with this
Open your own site on your phone, without scrolling, and work through this list. Each line is something the top quarter of Florida pool service sites does more often than the rest; the share after each line is how many sites in this study do it.
- A tap-to-call link on the first screen. Not the number as text: a link that dials. (54%)
- A button that says what happens next, such as "Get a quote" or "Book a visit", next to the call link. (57%)
- A call button that stays on screen while the visitor scrolls. (32%)
- A price or a starting price for weekly service, and whether chemicals are included. (24% show any price; 23% say clearly whether chemicals are included)
- Your license number, if you hold one; Florida requires it in contractors’ advertising [7]. (33%)
- What your insurance covers (general liability, workers’ comp), not just the word "insured". (8%)
- The towns or ZIP codes you serve, ideally on a page of their own. (29% have a separate page)
- What a weekly visit includes and how you report after each visit. (51% / 28%)
- Contract terms in one sentence: month to month or not, and how to cancel. (17%)
- How fast you start or call back. (32%)
- Reviews on your own site, and photos of the real people who come to the house. (59% / 32%)
Worksheet · your first screen
| On my phone, before scrolling, I can see … | Yes / no | Share of sites with it |
|---|---|---|
| My phone number | 54% | |
| A link that dials when tapped | 54% | |
| A quote, contact or booking button | 57% | |
| A call button that follows me when I scroll | 32% | |
| Nothing covering a fifth of the screen (pop-up, banner) | 82% |
Each "no" is a step a caller has to work around. Test on your own phone with mobile data, not office Wi-Fi.
Sources
- Google/Ipsos, Click to Call (Sep 2013). https://www.thinkwithgoogle.com/_qs/documents/1454/click-to-call_research-studies_hz2ytgv.pdf (accessed Oct 30, 2026)
- Invoca, Call Conversion Benchmarks Report: Consumer Services (2025; vendor). https://www.invoca.com/reports/the-invoca-call-conversion-benchmarks-report-consumer-services-2025 (accessed Oct 30, 2026)
- Oldroyd, McElheran, Elkington, "The Short Life of Online Sales Leads," Harvard Business Review (2011). https://scholarsarchive.byu.edu/facpub/9711 (accessed Oct 30, 2026)
- BrightLocal, Local Consumer Review Survey (2025). https://www.brightlocal.com/research/local-consumer-review-survey-2025/ (accessed Oct 30, 2026)
- Spiegel Research Center, How Online Reviews Influence Sales (2017). https://spiegel.medill.northwestern.edu/How-Online-Reviews-Influence-Sales (accessed Oct 30, 2026)
- Nielsen Norman Group, "Trustworthiness in Web Design" (1999, re-tested 2016). https://www.nngroup.com/articles/trustworthy-design/ (accessed Oct 30, 2026)
- The 2026 Florida Statutes, §489.119(5)(b) (license number in offers of services and advertisements) and §489.105(3)(l) (cleaning and water treatment need no license). http://www.leg.state.fl.us/statutes/ (Chapter 489, accessed Oct 30, 2026)
- Mohan, Buell, John, "Lifting the Veil: The Benefits of Cost Transparency," Marketing Science (2020). https://informs.org/News-Room/INFORMS-Releases/News-Releases/New-Research-Finds-Cost-Transparency-Can-Increase-Sales-20 (accessed Oct 30, 2026)
- ACHR News / ClearSeas Research, "Homeowners want prices listed online" (2023, trade survey). https://achrnews.com/articles/163905-homeowners-want-prices-listed-online (accessed Oct 30, 2026)
- Roofing Contractor, homeowner survey (2025, trade survey). https://roofingcontractor.com/articles/99160-homeowner-survey (accessed Oct 30, 2026)
- Deloitte for Google, Milliseconds Make Millions (2020). https://www2.deloitte.com/content/dam/Deloitte/ie/Documents/Consulting/Milliseconds_Make_Millions_report.pdf (accessed Oct 30, 2026)
- Portent, "Site Speed is (Still) Impacting Your Conversion Rate" (2022). https://www.portent.com/blog/analytics/research-site-speed-hurting-everyone.htm (accessed Oct 30, 2026)
- Think with Google, page load time statistics (2017). https://thinkwithgoogle.com/data/page-load-time-statistics (accessed Oct 30, 2026)
- web.dev, Core Web Vitals thresholds. https://web.dev/articles/vitals (accessed Oct 30, 2026)
- StatCounter, US platform market share (2026). https://gs.statcounter.com/platform-market-share/desktop-mobile/united-states-of-america (accessed Oct 30, 2026)
- Palm Beach County Consumer Affairs, "Swimming Pool Maintenance & Repair," Consumer Traps & Tips (file dated Feb 16, 2024). https://pbcdmzspptest.pbcgov.org/publicsafety/consumeraffairs/Tips/swimmingpoolmaintenance.pdf (accessed Oct 30, 2026)
- Ruberio crawl and index (this study), Oct 30, 2026: homepage and up to seven inner pages per site; Lighthouse 12.8 mobile, Chrome for Testing 153, median of 3 runs. Frames: Google Maps listings (Oct 27, 2026); Florida Division of Corporations (Oct 2026) and DBPR licenses (copy of Aug 10, 2026).
About this research
Who made this
Ruberio is a Dutch design studio: Ruben designs, Sven answers every email. We’re focused on websites for pool service companies and fence contractors in Florida, and wanted to know what a homeowner sees on a phone before saying anything about them; we found no published measurement of it. Data and scripts on request. Spot an error? Email hello@ruberio.com and we’ll correct it and note the change.
Want your own site with a tap-to-call link and a price on the first screen? We build a preview for $49 at us.ruberio.com; the $49 comes off the price if you keep it.
Cite this study
Ruberio Research (2026). 1 in 10 Florida pool service websites shows a phone visitor no phone number and no call, quote or contact button (No. 01). Data as of Oct 30, 2026. us.ruberio.com/research/florida-pool-service-website-conversion-index/
Questions
Does a missing call button mean a site loses customers?
This study doesn’t measure that. It measures readiness: what a site offers a visitor on a phone, not how many visitors call. The index weights rest on studies of retail, B2B, HVAC and roofing customers, not pool service.
How were the 114 websites chosen?
They come from Google Maps searches in four Florida metros (Tampa Bay, Orlando, Fort Lauderdale, Fort Myers), collected Oct 27, 2026, with national chains, franchises and directories removed. 167 had their own website; 114 loaded and were measured. The frame leans toward Tampa Bay (72 of 114).
Is 1 in 10 an exact figure?
No. It is 11 of 114 sites, with a 95% confidence interval of 5%–16%. Counting the 3 screens we could not judge as dead ends gives 12%. Sites were crawled on one day, Oct 30, 2026, and pop-ups and seasonal banners can change a first screen from day to day.
What should an owner check first?
Open your own site on your phone without scrolling. Look for a link that dials when tapped (54% of sites have one), a quote, contact or booking button (57%), and nothing covering a fifth of the screen. Test on mobile data, not office Wi-Fi.
