Measured, not claimed
How our sites actually perform
Every number here was measured on a simulated mid-range phone, against the live URL. Open Lighthouse and check any of them.
How these were measured
Lighthouse 12, mobile emulation — a mid-range Android phone, 4× CPU throttling, simulated slow 4G — median of three consecutive runs against the live URL. Measured August 8, 2026.
Mobile is the honest test. Desktop scores flatter almost every site; the same page routinely scores thirty points higher there. Most people find a local business on a phone, on a mediocre connection. That is the number that matters, so that is the number we publish.
What we build now
Seven sites we built from scratch to demonstrate the standard. These are fictional organisations, not clients — they exist so you can look at a finished build before you buy one.
Every one of them scores 100 on accessibility and 100 on best practices.
| Sample site | Speed | Accessibility | Content appears | Page weight |
|---|---|---|---|---|
| Turtle Creek Terrace Condominium association · Dallas, TX | 95 | 100 | 2.4s | 276 KB |
| Buffalo Bayou Condominium association · Houston, TX | 95 | 100 | 2.4s | 271 KB |
| Hill Country WCID No. 3 Water control & improvement district · Near Boerne, TX | 94 | 100 | 2.4s | 202 KB |
| Horizon Prep Academy Public charter school · Orlando, FL | 93 | 100 | 2.4s | 223 KB |
| Barton Springs Lofts Condominium association · Austin, TX | 94 | 100 | 2.4s | 432 KB |
| Lone Star Collegiate Academy Open-enrollment charter school · Austin, TX | 92 | 100 | 2.4s | 217 KB |
| Cypress Creek MUD No. 7 Municipal utility district · Greater Houston, TX The slowest of the seven. Trimming its ten font weights to five moved it from 76; the remaining gap is still being chased. | 86 | 100 | 2.4s | 288 KB |
Accessibility matters beyond good manners. Under ADA Title II, government and special-district sites must meet WCAG 2.1 AA — with a special-district deadline of April 26, 2028. Plain-language summary, not legal advice.
Real client rebuilds
The same measurement, applied to work we have actually shipped and to the sites we replaced.
ATX Monuments
Now
100
1.4s
Leon River Endodontics
Before
55
17.6s
Now
99
1.9s
Pembroke Cottage
Rebuilt again in August 2026 after measurement found it at 66 with a 28.3s LCP — slower on a phone than the site it replaced.
Before
56
17.9s
Now
98
2.3s
Peaceful Paws Pet Cemetery
Was 78. The hero was a 347 KB JPEG the browser could not discover until the stylesheet had parsed.
Now
95
2.9s
Stazon Roofing
Was 75. Eighteen images, none of them WebP, and a 268 KB hero declared halfway down the document.
Before
60
12.9s
Now
92
3.0s
What measuring found
In August 2026 we measured every site we had ever shipped, on a phone, for the first time. Before that we had been publishing estimates. Three of the five came back worse than we had claimed.
The worst was Pembroke Cottage at 66, with its main content taking 28 seconds to appear — slower on a phone than the Squarespace site it replaced. It was loading images straight from the client’s old host at full resolution, and a 416 KB icon library to draw two dozen icons. It now scores 98 on 508 KB instead of 5.6 MB.
Stazon Roofing and Peaceful Paws had the same disease in different forms: heroes the browser could not even discover until it had parsed the stylesheet, and not a single image in a modern format. They now score 92 and 95, up from 75 and 78.
One of those fixes silently did nothing for half a day. The stylesheet had changed but the theme version had not, so every visitor kept getting the cached copy while the corrected file sat on the server. We found it by measuring again after deploying, which is the only reason this page is not still reporting 80.
Why we publish this
A performance claim you cannot check is marketing. Every row above has a URL next to it — open Chrome’s Lighthouse panel, choose Mobile, and run it yourself. If a number here is ever wrong, we would rather you found it and told us than believed it.
Curious where your site lands?
Run the free checkup for an instant read, or ask for a full audit and we will measure it the same way we measured everything above.