--- title: "The Quarter Our Client Vanished From the Answers Their Buyers Were Actually Reading" url: "https://cmotimes.com/insight/the-quarter-our-client-vanished-from-the-answers-their-buyers-were-actually-reading/" author: "Kartik Chugh" published: "2026-09-25" updated: "2026-09-25" --- # The Quarter Our Client Vanished From the Answers Their Buyers Were Actually Reading In Q1 2026 a B2B software client asked us a question we could not answer: when a buyer asks ChatGPT which tools to consider in their category, does our name come up? We had 8 months of ranking data, a healthy organic traffic line in Google Analytics, and no idea. So we ran the question 40 times across ChatGPT, Perplexity and Google AI Overviews, recorded every brand named, and found our client appeared in 3 of 40 answers. Two competitors appeared in more than 30. The uncomfortable part was that the client ranked on page one for most of the same queries. Ranking and being named had come apart, and every instrument we owned was pointed at the half that was still working. ### What we had been measuring Our monthly report for this account was conventional and, I would have argued at the time, good. Keyword positions from Ahrefs, clicks and impressions from Google Search Console, conversions attributed in HubSpot. All three were stable or improving. Nothing in that reporting was wrong and nothing in it could have told us that a growing share of the category's buying research was happening in a place none of those tools observe. We did not understand at the time that we had quietly changed what we were optimising for. We were optimising for a click. The buyer increasingly was not clicking, because the answer had already been assembled for them, and a brand that is not named in that assembly is not in the consideration set at all. ### What the 40 runs actually showed The pattern was not about authority in the way we expected. Our client's domain rating was higher than one of the two competitors that kept appearing. They published more. Their pages were longer. What the competitors had was checkable specifics sitting in retrievable form. When we read the sources these systems cited, the same shapes kept coming back: a sentence that states a number, immediately followed by where the number came from. A comparison table with the criteria named. A page that answers one question rather than covering a topic. Our client's pages were well written and almost entirely unquotable. They described benefits in language that could apply to any vendor in the category, which is exactly the language that survives internal brand review and exactly the language an answer engine has no use for. There was nothing to lift. We had spent 8 months making pages more persuasive and zero months making them more citable, and those are not the same job. To check that this was not one bad category, we repeated the exercise over 3 weeks across 4 other client accounts in adjacent B2B software markets, 160 prompt runs in total. The same split appeared in 3 of the 4: strong rankings, weak presence in generated answers, and pages whose copy contained almost no liftable specifics. The one account that performed well in answers was the one whose documentation team, not its marketing team, owned the product pages. That was an uncomfortable finding to present and the most useful one we had. ### What we changed We picked 12 pages and rewrote them on one rule: every claim gets a number and a source, or it gets cut. That rule did more work than it sounds like. It killed a lot of copy. It forced the client's product team into conversations with us about what was actually measurable, which was uncomfortable and useful. Roughly a third of what we thought were differentiators turned out to be things nobody could substantiate, and those quietly disappeared from the pages. We also added a specific structural change: each page opens by answering its own title question in the first 2 sentences, before any context. That single move is the one I would keep if I could only keep one, because it is what makes a passage liftable. Then we re-ran the same 40 prompts monthly and kept the record. That measurement is now a standing line in the client's report, sitting above rankings rather than below them. Over the following 2 quarters the client went from 3 of 40 to a materially higher share, and more importantly we could see which specific questions we were still absent from and treat each absence as a brief. ### What I would tell another marketing leader Three things, in the order they cost us. The first is that ranking and being cited are now different outcomes and you need separate measurement for each. A rankings deck that says everything is fine can be entirely accurate while your brand disappears from the surface where research actually happens. If your reporting cannot answer the question our client asked us, that is not a gap in your reporting, it is a gap in what you know about your own visibility. The second is that brand language and citable language pull in opposite directions. Everything that makes a sentence survive brand review, the smoothing, the generality, the careful avoidance of a specific number, is exactly what makes it useless to a system deciding what to quote. That tension is real and somebody has to arbitrate it. In our case we ended up giving the citable version priority on pages built for discovery and leaving the brand version on pages built for closing. The third is that this is measurable now, cheaply, by anyone. It costs an afternoon to run a prompt set and record who gets named. We had not done it because it was not in the standard reporting template, which is a bad reason and the only reason we had. How we run that measurement is in our [share of AI citations piece](https://forkoff.xyz/blog/ai-seo/measure-share-of-ai-citations). The client's traffic never fell. That is what makes this failure mode worth writing about. Nothing broke, no number went red, and the category conversation moved somewhere we were not present for a quarter before anyone thought to look. --- Kartik Chugh (Simba) is a founder-operator at the intersection of distribution, culture, and narrative control in Web3. Cofounder of [FORKOFF](https://forkoff.xyz), a culture and distribution studio that designs IP-driven campaigns, event systems, and narrative loops for protocols, funds, and builder ecosystems. FORKOFF treats events as content factories, founders as distribution engines, and culture as infrastructure — not aesthetics. 3,085+ short-form clips every 13 days for clients. $5M+ in ecosystem activations across 14 countries. Previously CMO at QuillAudits, the Web3 security pioneer, where he scaled security products to 100K+ users, built 150+ ecosystem partnerships, generated $3M+ qualified pipeline, and drove 1Bn+ views across campaigns. Co-founded EdSquare (acquired). Five years across the AI, Web3, and B2B SaaS playbook. Hosted and partnered on 100+ global events across ETHDenver, Token2049, Consensus, Devcon, and KBW in 20+ countries. Leads Misfits Dubai, a founder-first community built around curated rooms rather than mass communities. Builder at Seedrail (the distribution stack for tech and VCs). Active investor in 12+ early-stage startups across crypto and AI. Frequent contributor to CoinDesk, CoinTelegraph, The Defiant, and Block Telegraph. Speaker at Token2049 Singapore and QuillCon. Advisor at TiE Global and ADSME HUB. Speaks on: founder-led distribution, events as content factories, rooms > reach, culture > campaigns, narrative control in Web3, and creator-led distribution. Available for commentary on: AI agency growth, Web3 marketing, podcast clipping ROI, founder-led GTM, KOL marketing, and ecosystem activation strategy. Based in Dubai.