AI Readiness Audit

    When a buyer asks ChatGPT who to call, are you in the answer?

    See who ChatGPT, Perplexity and Claude recommend instead of you. We ask twelve real buyer questions and show who gets named: you, the competitors you pick, or someone you have never heard of.

    Invite-only while in pilot. No code? Email marcus@grimmandco.org.

    What it checks

    The technical baseline

    • AI crawler access (GPTBot, ClaudeBot, PerplexityBot and others)
    • Content visible without JavaScript
    • Structured data that identifies you
    • Titles, descriptions and preview tags
    • Mobile page health (PageSpeed Insights)
    • Your homepage against up to four competitors

    What most audits leave out

    • Twelve buyer questions asked of ChatGPT, Perplexity and Claude
    • Whether you are cited, named, or absent, per engine
    • Who gets cited in your place
    • Which questions your pages answer today
    • Real AI referral traffic and landing pages (Google Analytics connected)

    Both lists in one report. The technical half is scored 0 to 100 on a transparent rubric; the citation half is counted, not scored.

    Why it matters

    Where AI traffic goes, and why it converts

    • AI visitors converted at roughly 4x the site average at an industrial manufacturer.
    • A wholesaler cited on zero category questions had the highest AI referral share, because ChatGPT names it for specific products.
    • AI traffic lands on product and service pages, not the homepage.

    How we checked this

    Both halves matter. The technical checks are the baseline your developers can act on this week. The content is where the work is, and where the buyers are.

    Sample report

    What you get

    Everything a technical AI audit checks, plus the part they leave out: whether you are in the answer. This is the top of a real report, anonymized.

    Sample report

    Simulation consultancy, run 2026-10-17

    Can AI read your site?

    Well prepared

    87 of 100

    You pass. 4 of 4 measured competitors pass too.

    2 Does AI recommend you?

    Cited in 2 of 12 buyer prompts

    #2 of 5 sites in your set

    Named without a link in 4 more answers.

    1 Technical scorecard

    Fix first: Article schema with datePublished and author, worth 2.7 points.

    • Crawler accessmeasured25 / 25
    • Extractabilitymeasured28.3 / 30
    • Structured datameasured9.9 / 15
    • Citation surfacemeasured13.7 / 15
    • Page healthmeasured10.4 / 15

    Top fixes

    • Article schema with datePublished and author

      +2.7 points, low effort, Dev

    • All JSON-LD blocks parse

      +2 points, low effort, Dev

    • LCP at or under 2.5 s

      +2 points, high effort, Dev

    Rubric published in full on every report.

    Citation share against your set

    • Competitor A4
    • Simulation consultancy2
    • Competitor B0
    • Competitor C0
    • Competitor D0

    3 Who is actually being cited

    • flexsim.com4 prompts
    • google.com4 prompts
    • anylogic.com3 prompts
    • en.wikipedia.org3 prompts
    • siemens.com3 prompts
    • thectoclub.com3 prompts

    The twelve buyer questions

    Cited: linked as a source. Named: mentioned in the answer without a link.

    1. 1.Best digital twin/virtual prototyping companies for complex system simulation?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    2. 2.How to choose simulation & optimization software?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    3. 3.List top supply chain optimization solutions for companies focused on cost cutting vs those focused on speed, how do choices differ?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    4. 4.Where can I buy digital twin simulation software that works across cloud and edge environments?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    5. 5.Where can I procure digital twin simulation software with strong system integration capabilities?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    6. 6.Which simulation platforms include expert technical support?

      Search ConsoleChatGPT: NoPerplexity: NoClaude: No
    7. 7.Who offers simulation consulting and digital twin development to improve manufacturing and supply chain operations?

      AI landing pageChatGPT: NoPerplexity: NoClaude: No
    8. 8.I need a consultant to debug and optimize our existing 3D discrete-event simulation model of a warehouse. Who can help?

      AI landing pageChatGPT: NamedPerplexity: NoClaude: No
    9. 9.Which simulation consulting firms have a team with expertise in industrial engineering, operations research and digital twins?

      AI landing pageChatGPT: NoPerplexity: NamedClaude: Cited
    10. 10.Which simulation consulting companies are hiring engineers to build manufacturing and logistics simulation models?

      AI landing pageChatGPT: NamedPerplexity: NoClaude: No
    11. 11.Which simulation consulting firms have recently expanded into Europe to support manufacturing and logistics clients?

      AI landing pageChatGPT: CitedPerplexity: CitedClaude: No
    12. 12.Which consulting firms can model our rail and intermodal network to help us evaluate terminal expansion and fleet investment?

      CategoryChatGPT: NoPerplexity: NoClaude: No
    1. 1

      Technical scorecard

      Scored 0 to 100 on a published rubric. Every gap shows the points it recovers, the effort, and who owns it. Hand the list to your developers.

    2. 2

      Does AI recommend you

      How often an AI assistant linked to you when a buyer asked about your category, ranked against the competitors you chose.

    3. 3

      Who is actually being cited

      The sites cited in your place. Often a directory, a publication or a firm you do not track. That is your real competition in AI answers.

    One run on October 17, 2026. AI answers change; each audit is a snapshot.

    How it works

    About two minutes, three steps

    1. 1

      We read your site

      We fetch your pages the way an AI crawler does, without JavaScript, and work out who you sell to.

    2. 2

      We set up the test

      You confirm the competitors you lose deals to, and we write twelve buyer questions with no brand names.

    3. 3

      We ask the engines

      ChatGPT, Perplexity and Claude answer each question, and we tally who is cited, named or left out.

    With Search Console and Google Analytics connected, the questions come from your real search queries, and the report adds your actual AI referral traffic and the pages it lands on.

    Before you run it

    Is AI recommending us, or our competitors?

    We ask twelve buyer questions across ChatGPT, Perplexity and Claude, then rank your citations against the competitors you choose.

    Keyword tool"best simulation software"

    This audit"We have a bottleneck in a plant and the spreadsheet model is not cutting it. Who do we call, and should we hire a consultancy or build this in-house?"

    Keyword tool"simulation consulting services"

    This audit"Our warehouse simulation model gives answers nobody trusts anymore. Who can audit and fix an existing discrete-event model?"

    Keyword tool"discrete event simulation companies"

    This audit"We are deciding whether to expand a distribution center or add a second shift. Which firms can model both options before we commit capital?"

    Example questions for a simulation consultancy selling to operations leaders.

    We read your site, classify your business model and who you sell to, and write the questions from that.

    Why are these questions better than a keyword list?

    Built by a working fractional CMO who sits in the meetings where buyers move from a problem to a purchase, the method was tested on real client analytics. That is why the questions sound like a buyer, not a keyword list.

    Will AI even crawl my site?

    The technical baseline covers crawler access, JavaScript-free content, structured data and page health, scored 0 to 100 on a published rubric. Every gap shows recoverable points, effort and owner.

    Is this just another dashboard?

    One run, one report, both halves, nothing to configure and no subscription. The method was checked against real AI referral traffic.

    What this does not measure

    Off-site mentions, the full quality of your content, and the complete set of questions buyers ask. The citation check samples twelve prompts; it is a signal, not a census. A readable site is the baseline, not the finish.

    After the report

    What happens next

    If the baseline has gaps

    The report lists the fixes in priority order, with effort and owner. Most teams can clear them in-house within days.

    If the baseline is clean and you are not cited

    That is content work, the harder half. The Content Priority Audit is the next step.

    Run the audit
    Marcus Grimm

    Built by Marcus Grimm, fractional CMO for complex B2B.