Hawaii Small-Business Websites on a Phone
We loaded the home pages of 553 Hawaii small businesses on a phone. In Google’s lab test, 88% took longer than 2.5 seconds to show their main content, and the typical page took 7.1 seconds.
Most pages are slow to show their main content.
Largest Contentful Paint (LCP) is the moment the biggest thing on the first screen, usually a photo or a headline, finishes drawing. Google counts 2.5 seconds or less as good. In the mobile lab test, 453 of 517 home pages (88%) took longer. The typical page took 7.1 seconds, and restaurant and food-truck pages were slowest, with a median of 9.4 seconds.
R2 Median time to main content, seconds. The line marks Google’s 2.5 s.
- 6.4 sTradesn 140
- 7.0 sCleaning & pestn 118
- 6.6 sAuto & personaln 129
- 9.4 sFoodn 130
- 7.1 sAll sitesn 517
More than nine in ten show no rating where a visitor first looks.
We looked for a star rating or a review count in the first screen, with no scrolling. 457 of 485 sites (94%) showed none. Even among the top-rated quarter of businesses on Google, 88% showed none, against 96% of the rest.
R4 No rating in the first screen, share of home pages
- 93%Trades121/130
- 93%Cleaning & pest104/112
- 96%Auto & personal115/120
- 95%Food117/123
- 94%All sites457/485
Half show no phone number in the first screen.
A phone number counted when it was a tap-to-call link or visible text in the first screen, with no scrolling. 251 of 485 sites (52%) had none there. Restaurants and food trucks were furthest behind at 78%; home and building trades did best at 40%. A number inside a closed menu or a picture doesn’t count here.
R5 No phone number in the first screen, share of home pages
- 40%Trades52/130
- 43%Cleaning & pest48/112
- 46%Auto & personal55/120
- 78%Food96/123
- 52%All sites251/485
Six in ten have no form anywhere on the home page.
No contact, quote or booking form at all, counting forms in pop-ups and embedded widgets: 289 of 485 sites (60%). Auto repair and personal services was highest at 73%. A form field in the first screen was rarer still: 6%, 27 of 485.
R6 No form on the home page, share of home pages
- 51%Trades66/130
- 50%Cleaning & pest56/112
- 73%Auto & personal88/120
- 64%Food79/123
- 60%All sites289/485
Better-reviewed businesses’ sites were no faster.
The top-rated quarter’s home pages had a median LCP of 7.6 seconds, against 6.8 for the rest. 26% scored under 50 for performance, against 22%. The 95% ranges overlap (20–34% and 18–26%), so we read this as no difference: a good reputation on Google and a fast site didn’t go together in this sample.
R2 Median time to main content, seconds
- 7.6 sTop-rated quartern 133
- 6.8 sThe restn 384
R3 Performance score under 50
- 26%Top-rated quarter35/133
- 22%The rest83/384
Fig. 06 · Rows R2 and R3 split by Google rating. Top-rated quarter: the top 25% of each trade by rating weighted for review count. The 95% ranges overlap (N4).
What it means for your site.
The table’s rows, turned into four checks you can run yourself with your own phone and a browser.
- R5
Your number, first screen
Open your home page on your phone. Is your number on the first screen, as a link you can tap?
- R4
Your rating, in view
Can a visitor see your Google rating or review count without scrolling?
- R6
A way to write to you
Is there a form on the home page for a quote, a booking or a question?
- R1
Your speed, measured
Run your home page through PageSpeed Insights on mobile. Largest Contentful Paint at 2.5 seconds or less is Google’s “good” mark.
Method and limits, as notes
The results table cites these notes by number. Together they are the whole method: how the sample was drawn, what was measured, and what the numbers can’t say.
- N1
Sample. Google Places text searches for 14 business categories in five towns on four islands (Honolulu, Kahului, Hilo, Kailua-Kona, Lihue), 8 November 2026. Places that list their own website, with marketplace and social pages, chains and duplicate domains removed. 577 selected; 24 set aside after review (15 third-party listing pages, 9 chains); 553 home pages.
- N2
Speed. PageSpeed Insights (Lighthouse 13.6.0), mobile, one run per home page; 517 measured. LCP is Largest Contentful Paint: when the biggest thing on the first screen finishes drawing.
- N3
Page checks. Our own visit in Chrome at 390 × 844 pixels, one page per site, no scrolling, robots.txt honoured; 485 measured. Not measured: 10 excluded by robots.txt, 8 bot-blocked, 34 unreachable, 11 error pages, 3 not a business site, 2 navigation errors.
- N4
Thresholds and ranges. LCP over 2.5 s is Google’s “good” mark; a performance score under 50 is Lighthouse’s red band; the first screen is the top 390 × 844 pixels. Each percentage has a 95% range; where two groups’ ranges overlap, we don’t treat the difference as real.
- N5
Rating, defined. A rating counts when a star rating or a review count is visible in the first screen, with no scrolling: 94% showed none (457 of 485). Counting rating data in the page’s code as well, 90% had none (437 of 485).
- N6
Top-rated quarter. The top 25% of each trade by Google rating weighted for review count: 138 of 553 home pages.
- N7
Limits. PageSpeed Insights is a lab test on an emulated phone, run once per site; real visitors can see faster or slower pages. The home page may not be the page an ad sends people to. The sample follows Google’s ranking, not a random draw, and covers only businesses that list a website. Stars drawn as images and numbers inside pictures are missed; “first screen” is geometric. Some chains and franchises may remain. One snapshot, November 2026.
Questions about the study
How the sample was built, what the numbers can and can’t say, and how to check your own site.
How were the businesses chosen?
Google Places searches for 14 categories in five towns across Oʻahu, Maui, Hawaiʻi Island and Kauaʻi. We kept businesses that list their own website, set aside marketplace pages, chains and duplicates, and measured 553 home pages.
Is this what real customers see?
Close, but not identical. Speed comes from Google’s lab test on an emulated phone, run once per site; real visitors on real networks can see faster or slower pages. The other checks come from one visit in Chrome at a phone’s screen size.
Do you publish results for individual businesses?
No. Names and web addresses stay in a private dataset; this page reports totals only.
Why only the home page?
It’s the page Google lists for each business. Ads often send people to other pages, which this study did not measure.
Can I check my own site?
Yes. Run your home page through PageSpeed Insights on mobile for speed, and look at the first screen on your own phone for the rest. Or send it to us through the contact form.
Check Your Own Site
Send us your website through the contact form and we’ll look at the same things this study measured: speed on a phone, your number, your rating and a way to reach you.
It’s free, and we reply within one business day.
Or get the free check of your site
For the wider picture of how Hawaii small businesses market themselves, see our Hawaii Small Business Marketing Report 2026.