US law firm websites: practices, Spanish, intake

What does a record look like?

• California: 1,145 law firms found (1,145 with a website read, 0 from official registries only); coverage: 5% of the 22,813 employer law offices the Census Bureau counts (County Business Patterns 2023) and about 2% of the estimated 66,725 law firms with a website; a sample; of the websites read, 15.4% show Spanish-language pages, 23.1% show a 24/7 intake statement, 41.7% show legal structured data, 56.2% show a free consultation statement, 27.2% show a no-fee-unless-we-win or contingency statement, 5.5% show a chat tool, 4.5% show an online booking tool, 18.7% show review markup (schema.org Review or AggregateRating), 32.7% show an llms.txt file.

  Status P · checked October 28, 2026 · source: US law firm websites: the firms' own websites, read up to 2026-10-28

• Texas: 12,444 law firms found (883 with a website read, 11,561 from official registries only); coverage: 90% of the 13,751 employer law offices the Census Bureau counts (County Business Patterns 2023) and about 2% of the estimated 51,690 law firms with a website; a census; of the websites read, 17.0% show Spanish-language pages, 24.8% show a 24/7 intake statement, 46.2% show legal structured data, 55.6% show a free consultation statement, 24.8% show a no-fee-unless-we-win or contingency statement, 5.5% show a chat tool, 6.9% show an online booking tool, 19.1% show review markup (schema.org Review or AggregateRating), 38.3% show an llms.txt file.

  Status P · checked October 28, 2026 · source: US law firm websites: the firms' own websites, read up to 2026-10-28

• Florida: 820 law firms found (820 with a website read, 0 from official registries only); coverage: 5% of the 17,243 employer law offices the Census Bureau counts (County Business Patterns 2023) and about 2% of the estimated 47,699 law firms with a website; a sample; of the websites read, 17.7% show Spanish-language pages, 28.3% show a 24/7 intake statement, 42.7% show legal structured data, 57.8% show a free consultation statement, 24.5% show a no-fee-unless-we-win or contingency statement, 5.1% show a chat tool, 5.5% show an online booking tool, 17.7% show review markup (schema.org Review or AggregateRating), 39.8% show an llms.txt file.

  Status P · checked October 28, 2026 · source: US law firm websites: the firms' own websites, read up to 2026-10-28

The full records add the dates, the quoted source text and every field in the CSV and JSON files below.

What are the limits?

• Coverage is shown two ways on every row: (a) the firm websites we read against our estimate of the law firms with a website, and (b) all firms found, with or without a website, against the employer law offices the Census Bureau counts (the latest County Business Patterns). A state is called a census only when (b) reaches seven in ten; until then its rows are a sample. Compare shares, not totals, where coverage is low.
• Firms and establishments are not the same unit: a firm can have several offices, and state registries include solo practices without employees that County Business Patterns leaves out, so (b) can pass 100%. The website estimate in (a) is a capture-recapture estimate from two independent site lists and stays provisional until our first full pass over the Common Crawl list ends.
• Licences differ. These aggregate rows are CC BY 4.0. The firm-level rows (also shown on the /firms/ pages) (one per firm, with the page each fact was read on) are a separate download (listed in /data/firm-web-census/firms/index.json, explained in /data/firm-web-census/README.txt) under the Open Database License (ODbL), because the list of sites comes partly from OpenStreetMap (© OpenStreetMap contributors): if you publish an adapted version of those rows, it must stay under the ODbL.
• Registry firms come from official open-data portals (New York's attorney registrations and business entities; the business registries of Colorado, Oregon, Connecticut, Texas and Pennsylvania), selected by law-firm words in the registered name. For them we show only what the registry publishes: the name as registered and the office's city, ZIP code and county. Other states publish no such open file yet, and bar directories are never used.
• Facts are read by fixed rules from up to a dozen pages per firm. A practice the firm handles but never names on those pages is missed, and a word can be read as a practice the firm does not take. A statement such as 'available 24/7' or 'free consultation' is recorded as the firm states it; we do not test it.
• Sites that block automated visits, ask for a challenge, disallow our crawler in robots.txt or do not answer are not read. A firm is placed in the counties of the office addresses on its own pages (ZIP code to county with HUD's USPS crosswalk).
• County rows with fewer than 3 firms show no count, county-and-practice cells with fewer than 3 firms are left out, and shares and medians need at least 5 firms.
• The rows describe what websites say on the date read, not the quality of a firm or a lawyer, and no individual lawyer's details are stored. A firm can ask for a correction or to be left out through the data corrections page; we act within 7 days.
• Weekly change (firms new to the sample and firms that stopped answering) starts once two weekly runs exist; monthly snapshots of every row are kept so change over time can be measured.
• Records older than 21 days since their last check are marked stale in the files until we re-check them. Every confirmed error we fix is listed in the data corrections log.

Found a wrong value? Report an error with the record id and the source that shows the right value.

Running a firm? You don’t need the raw data: we handle your state’s ad rules for you. See where your firm stands first.