After-Hours Contact and Reply Times on Texas PI Firm Websites (2026 Study)

    Nimbus Marlowe research
    Last updated November 7, 2026
    7 min read

    We read the public websites of 200 small plaintiff personal injury firms in Texas between 30 October and 1 November 2026. 132 of them (66%) stated a way to reach the firm outside business hours. 19 (9.5%) stated how fast the firm replies to a new inquiry. Nobody at any firm was called, emailed or contacted, and this page names no firm.

    Published 6 November 2026 by Nimbus Marlowe research. Sites read 30 October to 1 November 2026. Next update: May 2027.

    What did the study find?

    • •66% state an after-hours route (132 of 200; 95% interval 59.2% to 72.2%). Defined before collection as any of three things: a phone line the site describes as 24/7, after hours, nights and weekends or similar; a live chat widget; or a way to text the firm.
    • •53.5% state phone availability outside business hours (107 of 200).
    • •9.5% state a specific reply time for a new inquiry (19 of 200; 95% interval 6.2% to 14.4%). Specific means a number of minutes, hours or days, "within the hour", "same day" or "today".
    • •10.5% make only a general promise (21 of 200), such as "as soon as possible" or "promptly". This count carries lower confidence, explained in the reliability section below.
    • •93.5% have an online contact or intake form (187 of 200).

    So someone looking at 9pm will usually read that the firm can be reached, and will rarely read when to expect an answer. Clio's 2019 Legal Trends Report found that 79 percent of potential clients expect a response within 24 hours of contacting a law firm.

    How do the numbers change with city size?

    The 24 cities sit in three groups of eight: large metro (Houston, Dallas, Fort Worth, San Antonio, Austin, Arlington, Plano, El Paso), mid-size (Corpus Christi, Lubbock, Laredo, McAllen, Waco, Beaumont, Amarillo, Tyler) and smaller cities (Brownsville, Abilene, Odessa, Midland, Longview, College Station, San Angelo, Wichita Falls).

    • •An after-hours route (phone, chat or text): 66% of all 200; large metro 79.5%, mid-size 61.8%, smaller cities 45.5%.
    • •Phone availability outside business hours: 53.5% of all 200; large metro 69.3%, mid-size 50%, smaller cities 27.3%.
    • •A live chat widget: 31% of all 200; large metro 37.5%, mid-size 27.9%, smaller cities 22.7%.
    • •A text or SMS option: 12% of all 200; large metro 15.9%, mid-size 5.9%, smaller cities 13.6%.
    • •An online contact or intake form: 93.5% of all 200; large metro 94.3%, mid-size 94.1%, smaller cities 90.9%.
    • •A specific reply time for a new inquiry: 9.5% of all 200; large metro 11.4%, mid-size 8.8%, smaller cities 6.8%.
    • •Only a general reply promise: 10.5% of all 200; large metro 17%, mid-size 4.4%, smaller cities 6.8%.
    • •Online scheduling: 4.5% of all 200; large metro 6.8%, mid-size 1.5%, smaller cities 4.5%.
    • •A Spanish-language option: 59% of all 200; large metro 65.9%, mid-size 55.9%, smaller cities 50%.
    • •After-hours phone, with no chat or text: 27% of all 200; large metro 30.7%, mid-size 30.9%, smaller cities 13.6%.
    • •An email link: 33% of all 200; large metro 31.8%, mid-size 26.5%, smaller cities 45.5%.
    • •A client login or portal link: 7.5% of all 200; large metro 12.5%, mid-size 2.9%, smaller cities 4.5%.
    • •Google Maps listing open 24 hours every day: 61.5% of all 200; large metro 87.5%, mid-size 50%, smaller cities 27.3%.

    Large-metro sites were the most likely to state an after-hours route, and smaller-city sites the least likely. Reply-time statements stayed rare at every size, between 6.8% and 11.4%. The smaller-city group has 44 firms, so its intervals are wide: 31.7% to 59.9% for the after-hours route.

    Spanish stands out: 59% of sites offer a Spanish page, a language switch, or a line saying staff speak Spanish.

    How was the study done?

    • •Rules first. The question, the sample, every coding rule and the analysis were written and committed before any firm's site was read (commit a479ecf, 30 October 2026). Every change after that is listed below.
    • •Where the firms came from. Google Maps results for "personal injury lawyer", "car accident lawyer" and "injury attorney" in each of the 24 cities. The smaller cities came up short, so a second pre-planned wave searched "truck accident lawyer" and "personal injury attorney" too.
    • •Who counted. 444 firm websites were screened. A firm stayed in when it had its own website, personal injury was its main practice, it worked the plaintiff side, all its offices were in Texas, it was a law firm and not a referral service, and its site showed 1 to 10 attorneys. 230 passed. The first 200 in a fixed random order became the sample. Only 44 smaller-city firms passed, so the other groups filled the gap in the same order.
    • •What was read. A browser loaded each home page, waited for chat widgets, and opened up to three more pages: contact, team or about, and a free consultation page. Pages blocked by robots.txt were skipped.
    • •How it was coded. Phone hours, chat, text, forms, scheduling and Spanish were coded from page text and page code. For reply time, a wide net pulled every sentence pairing a speed word with a contact word, and an AI reader (Claude) judged each one as specific, general or not a reply promise. It judged 631 sentences. Testimonials, lines about insurers, promises to existing clients and invitations like "call today" do not count.
    • •Never collected: reviews and star ratings.
    • •Analysis: unweighted shares with Wilson 95% intervals.

    Changes after the rules were committed

    1. A robots.txt parsing bug was fixed and the one site it skipped was read.
    2. The second search wave ran, as the rules required when the sample came up short.
    3. A sampling bug was fixed. The rule itself did not change.
    4. 17 firms the insurance-defence screen had wrongly excluded were put back, because the text described a lawyer's past work. Two firms listing insurance defence as a current practice stayed out.
    5. The agreement check below was done by a blind AI coder. After the first two checks, the reply-time, Spanish and chat rules were rewritten and all 200 sites were recoded.

    How reliable is the coding?

    A second coder, a separate Claude instance that never saw the first coding, coded 60 sites from fresh screenshots and page text, in three sets of 20. The first two sets shaped the rule changes above. The third shaped nothing, so it is the clean measure of the final rules.

    • •Phone outside business hours: 95% agreement, kappa 0.90.
    • •Live chat: 90% agreement, kappa 0.79.
    • •Text option: 90% agreement, kappa 0.00.
    • •Online form: 95% agreement, kappa 0.64.
    • •Reply time, specific or not: 95% agreement, kappa 0.64.
    • •Reply time, three levels: 80% agreement, kappa 0.52.
    • •Online scheduling: 95% agreement, kappa 0.00.
    • •Spanish option: 95% agreement, kappa 0.87.

    The headline reply-time split agreed on 58 of 59 sites across all three sets. The three-level code fell short of the 90% bar, which is why the general-promise count never leads. The kappas of zero come from sets with almost no positives: the second coder counted a WhatsApp button, a "text the attorney" button and one scheduling tool the first coding missed.

    What can't this study tell you?

    • •How any firm behaves. Every number describes what a website says. A firm that states no reply time may answer in minutes, and a stated 24/7 line tells you nothing about who picks up.
    • •Firms outside Google Maps results. Maps favours firms that invest in local visibility.
    • •Anything outside Texas, or any change over time. One state, one snapshot.
    • •Every chat widget. Widgets that load late, or come from vendors outside the study's list, are undercounted.
    • •Promises written in Spanish, or on other pages. Reply times were found in English only, and only four page types were read.
    • •Firms whose home page never says injury, accident or wreck. They were screened out, including one titled only in Spanish.
    • •Current attorney counts. Team pages can lag.
    • •A human second opinion. Both readers were AI.

    Download the aggregate counts (CSV, no firm names or websites).

    This study was run by Nimbus Marlowe, the AI operations layer for personal-injury law firms. Questions about the method: [email protected].

    Frequently asked questions

    Does this mean most Texas PI firms reply slowly?

    The study can't say. It reads what sites say about themselves, and a firm can reply fast without ever printing a time.

    Was any firm contacted?

    No. No calls, forms, chats or emails. Clio's secret-shopper studies called firms. This one only read public websites and Google Maps listings.

    Which firms were in the study?

    The sample is not published, and no firm is named, rated or ranked. Only the totals on this page and in the download leave the study.

    What counts as a specific reply time?

    A stated number of minutes, hours or days for answering a new inquiry, or "within the hour", "same day" or "today" tied to a reply. "We'll respond promptly" counts as general.

    How was AI used?

    A Claude reader judged each candidate reply-time sentence, part of the Spanish coding and some chat widgets. A second, separate Claude coder checked 60 sites without seeing those results.

    When will it be updated?

    May 2027, with the same rules, so the change between the two dates becomes the next finding.

    See it handle a live call.

    Contact us

    Tell us what you are trying to automate and we will get back to you.