Most law firm website audits dwell on Lighthouse scores or server response times while ignoring the actual content of the HTML document itself. How much of that document is useful text? How much is inline CSS that could be externalized and cached? How many images load immediately even though the visitor will never scroll to them? We analyzed the raw HTML source of 303 law firm homepages across 25 U.S. markets to answer these questions with data rather than assumption.

Methodology

We collected the homepages of 303 law firms identified through Google organic search results across 25 high-growth, mid-size U.S. markets and two practice areas (personal injury and family law). For each homepage we fetched the raw HTML source using a headless browser and parsed it to extract the following.

  • HTML document size: total bytes of the raw HTML response.
  • Inline CSS: total bytes within <style> tags embedded in the document.
  • Inline JavaScript: total bytes within <script> tags, excluding external src references.
  • External resources: a count of external stylesheets, scripts, images, and other linked resources.
  • Image optimization signals: use of loading="lazy", srcset attributes, and <picture> elements.
  • Content metrics: word count, text-to-HTML ratio, heading count, paragraph count, and link count.

All data was collected in a single pass. Sites that returned redirect chains, error pages, or empty responses were excluded from the analysis.

The numbers at a glance

Headline figures across 303 homepages
MetricValue
Average HTML size262KB
Average external resources48.7
Average text-to-HTML ratio5.3%
Sites using lazy loading28.4%

HTML size distribution

The distribution of HTML document sizes across all 303 sites shows that most law firm homepages fall in the 100 to 250KB range. Nearly one in ten sites exceeds 500KB of raw HTML, a threshold that creates noticeable delays on mobile connections.

Homepages by HTML document size
HTML size rangeSitesShare
Under 100KB8829.0%
100 to 250KB11838.9%
250 to 500KB6421.1%
500KB to 1MB268.6%
Over 1MB72.3%

The 100 to 250KB bucket is the most common, containing 38.9 percent of all sites. This is the default weight for a typical CMS-built law firm homepage. The under-100KB group (29 percent) tends to include either well-optimized custom builds or very minimal single-page sites. The seven sites exceeding 1MB of HTML alone represent extreme cases of template bloat, inline SVG abuse, or massive inline script payloads.

The inline resource problem

One of the biggest contributors to HTML weight is code that has been embedded directly into the document rather than loaded as external, cacheable files. Across our sample:

  • Inline CSS averages 72KB per page (median 27KB, max 896KB). This CSS is re-downloaded on every single page load because it lives inside the HTML document rather than in a separate cached stylesheet.
  • Inline JavaScript averages 59KB per page (median 20KB, max 5MB). The worst offenders embed entire analytics libraries, chat widgets, and tracking scripts directly in the HTML.

Combined, the average law firm homepage carries 131KB of inline CSS and JavaScript. That is half the total HTML weight given over to code that could be externalized into separate files and cached by the browser indefinitely.

The gap between average and median tells the real story. The median inline CSS (27KB) and median inline JS (20KB) are far below the averages, which means a relatively small number of extremely bloated sites drag the averages up. For the typical homepage, inline resources account for roughly 47KB, still significant but manageable. For the outliers, inline code alone can exceed the entire HTML weight of an optimized site.

The image optimization gap

The average law firm homepage loads 30.6 images. At that volume, how those images are delivered has an outsized effect on performance.

Lazy loading adoption

Only 28.4 percent of the 303 sites use the loading="lazy" attribute on any of their images. Among those that do, an average of just 20.2 percent of images on the page are marked as lazy. Even the sites that have adopted lazy loading apply it to only a fraction of their images.

The practical impact: on a page with 30 images and no lazy loading, the browser attempts to download all 30 at once on page load, whether or not the visitor will ever scroll below the fold. That directly hurts Largest Contentful Paint and overall Time to Interactive.

Responsive images

35.6 percent of sites use the srcset attribute to serve different image sizes by viewport width. That is a modest improvement over lazy loading adoption, but it still means nearly two-thirds of law firm sites serve the same image file to a phone that they serve to a desktop monitor.

The combination is telling: most law firm homepages load too many images, load them all at once, and serve desktop-sized files to mobile devices. These are not obscure optimizations. Lazy loading and srcset are native HTML features supported by every modern browser.

Content density

Beyond performance, page weight reveals how much of a homepage is actual content versus structural code.

Word count

The average law firm homepage carries 1,677 words (median 1,310). That is within the range SEO research generally considers adequate for service pages, though the spread is enormous, from sites with nearly zero visible text to one homepage with over 23,600 words.

Text-to-HTML ratio

The average text-to-HTML ratio is 5.3 percent (median 4.5 percent). For every 100 bytes of HTML delivered to the browser, fewer than 6 bytes are actual readable text. The remaining 94-plus bytes are markup, inline styles, inline scripts, and structural code.

A low text-to-HTML ratio is not inherently bad; modern web development requires a certain amount of structural markup. But combined with the inline resource data above, it describes homepages carrying significant dead weight. A 262KB HTML document with a 5.3 percent text ratio contains roughly 14KB of readable content wrapped in 248KB of code.

For context, the maximum text-to-HTML ratio in our sample was 33.8 percent, which demonstrates that it is entirely possible to build content-rich law firm homepages without the code overhead.

Limitations

This study carries several important caveats.

  • HTML only. We measured raw HTML document weight, not total page weight (which would include all external CSS, JavaScript, images, fonts, and other assets). Total page weight is typically 5 to 20 times larger than HTML weight alone.
  • Homepage only. We analyzed only the homepage of each site. Interior pages, practice area pages, attorney bios, and blog posts, may have very different weight profiles.
  • Point-in-time snapshot. All data was collected in a single pass. Sites are updated frequently, and weights can change with any content edit or plugin update.
  • No rendering analysis. We measured source code metrics, not rendering performance. A heavy page with effective caching may still perform well for repeat visitors.
  • CMS variability. We did not control for CMS platform. WordPress sites with heavy themes, Wix sites, custom builds, and directory listings all appear in the same dataset.

Full methodology details

Data collection

Sites were identified through Google organic search results across 25 high-growth, mid-size U.S. markets using queries for personal injury and family law firms. We fetched each homepage using a headless browser (Puppeteer) with JavaScript enabled, capturing the full rendered HTML source after initial page load.

Fields extracted per site

Domain, URL, HTML byte count, inline CSS byte count (sum of all <style> tags), inline JavaScript byte count (sum of all inline <script> tags), external stylesheet count, external script count, total external resource count, image count, lazy-loaded image count, srcset image count, <picture> element count, visible text word count, text byte count, text-to-HTML ratio, heading count, paragraph count, and link count.

Calculations

Text-to-HTML ratio: (text bytes / HTML bytes) x 100. Lazy load percentage: (images with loading="lazy" / total images) x 100, calculated only for sites with at least one image. All averages are arithmetic means. Medians are reported alongside averages where the distribution is skewed.

Sample

303 unique law firm homepages across 25 markets: Georgetown TX, Wilmington NC, Myrtle Beach SC, Murfreesboro TN, Concord NC, Huntsville AL, Port St. Lucie FL, Naples FL, Daphne AL, Cary NC, Round Rock TX, New Braunfels TX, Goodyear AZ, McKinney TX, St. George UT, Provo UT, Fort Collins CO, Bend OR, Castle Rock CO, Roseville CA, Carmel IN, Sioux Falls SD, Fargo ND, Rochester MN, Coeur d'Alene ID.

The reason this matters for our own clients is direct. The Nitrosite Standard exists to invert nearly every figure above: per-page CSS inlined only where it earns its place, no uncacheable script payloads, edge-deployed static files, and a Lighthouse score of 100 against an industry whose average is 34. The data here is the problem; the architecture is the answer.