Skip to content
    Case study 03. Regulated infrastructure

    Being the answer an AI model gives now matters more than a high ranking on Google.

    Anonymized client. Account attested by Average Robot.

    The client is not named and the account reports no figures. It is Average Robot's own record of the engagement.

    1. The tension

      The way people find information is changing under everyone at once, and the bigger you are, the more exposed you are. The page you spent a decade getting to rank is not the answer an AI model now hands back in a sentence. A large, regulated company has the most content to maintain and is the slowest to turn when search changes.

    2. The surface story

      A large, regulated infrastructure company running the standard corporate search playbook: a complex message, a fragmented content strategy, and years of web pages built for a search engine that rewarded volume and keywords.

    3. The default path

      Keep pouring maintenance into a sprawling estate of pages optimized for a version of search that is on its way out, and find out too late that the budget was buying less every quarter.

    4. The ground truth

      Working through People, Process, and Platform, we found the effort and the budget were both there. The whole content estate was built for a version of search that is fading, so every quarter of maintenance returned less than the one before. The work was sound. The way search works underneath it had changed, and no one inside had the remit to say so out loud.

    5. The reframe

      The new goal was to be the answer an AI model gives.

    6. The design

      We showed them where search is heading, then set out an approach built on leaner pages and content written to answer the questions people actually ask, rather than telling the product story the old way.

    The outcome, across People, Process, and Platform.

    Process

    The approach moves a fragmented content estate toward a leaner, answer-led approach, with pages written around real questions instead of keyword volume.

    Platform

    It reorients the content for how AI models surface answers, so the company is set up to be the answer rather than buried under its own back catalog.

    Being the answer an AI model gives now matters more than a high ranking on Google.

    Without people, AI is just average.

    How we work

    • How we help: the framework behind every case, People before Process before Platform.
    • Our products: the ways to work with us, starting with the free Assessment.
    Weekly newsletter

    With People

    Every Tuesday: the people side of making AI pay.

    • One story from the week's news about AI at work, read through our book, The Elephant in the Algorithm.
    • A few links worth your time, and most weeks a podcast conversation.
    Privacy Policy