Adgully Bureau
| 3 hours ago
Views: 41
Your browser does not support the audio element.
By: Ambika Sharma, Founder, Chief Strategist at Pulp Strategy and Product Architect of NeuroRank
Roger Lynch runs Condé Nast. Last year he told his teams to plan every business as if search traffic were zero. He was not being dramatic. Three years running, the company’s own search forecasts had been too optimistic, and the traffic kept falling faster than the plan.
Look at what he did. He stopped counting the click, and what arrives in its place is a sentence.
Everything in marketing was built for the click. Ranking, position, share of search, the page. Your buyer asks a model which brand to buy and gets none of them. She gets one paragraph with three names, and either yours is one of the three or it is over before you knew it began.
Addition stopped working. For twenty years the job was to add, more reach, more spend, more content, and the brand that put in more climbed higher. AI search does not add. It selects. It returns two names, sometimes four, and there is no second page. You are on that list, or you are out of the decision.
And you do not write it. McKinsey put a brand’s own website at 5 to 10 percent of what an AI model uses to describe it. The other 90 percent is other people’s writing, and it fails four ways. Omitted. Replaced by a competitor. Hallucinated into something you are not. Zero Leads back to you. Fixing that is not a campaign. It is a change in how marketing is built and run. Here is what changes.
Marketing, corporate communications, PR, and the website team have always run independently. Each owns its own surface, on its own scorecard. That worked when the surfaces were separate. NeuroRank covers recommendations that land with each of these teams already. AI search reads all of them and speaks one description of the brand, so the four now have to work as one. This is a change to the structure, not the content calendar, teams need to come together and unify in action, and in most large companies it has not happened yet. Start here.
Somewhere in the last decade, teams stopped reading what the SEO team published. As long as the backlinks grew, nobody checked the words. Those words are training data now. Old specs, retired category lines, claims written for a crawler, the model quotes all of it as current fact. Audit the archive and fix it. And treat it as a legal document, because the courts have started to. In late May a Munich court stopped Google over AI Overview claims about two publishers that sat in none of the sources cited. In the United States, the courts have so far leaned the other way and protected the AI maker, on the ground that disclaimers mean no reasonable reader treats the output as fact. We know that a lot of consumers do, data says 68% or more users believe AI answers blindly. At some point the litigation will move to the source of the error.
Build your own AI search governance
Stop guessing. The brand is the entity at stake here, undoing a badly done GEO practice is close to impossible. Ownership within the team and the tech to keep track are the very least of the prerequisites. You cannot manage what you cannot see, and most brands have never seen what the models say about them at scale. Build a permanent view of it, yours, not the agency’s, not a one-off audit. Monitor the four engines, own the description, drive the fixes whoever executes them. This is the discipline NeuroRank was built to run. Governance is yours to hold. Execution can sit anywhere.
Trial and error is not free here. Every untracked content push, every guess at what moved the answer, spends real budget against a target that resets monthly. Spray and pray was always wasteful. Here it is also invisible, landing on no dashboard you currently own. Name the queries, give every fix an owner, measure the lift, repeat. This is a time bound continuous discipline. That is one thing it has in common with SEO.
The model will state things about your brand that are untrue, delivered with complete composure. You are responsible for correcting them, whether or not you caused them. The impact on consumers engagement with your brand is massive, it’s also irreversible. From a legal standpoint AI-prompted litigation is not coming. It is here. The German court ruling is the first, not the last.
Paid media is now tied to AI search
Bad generative search optimization bleeds your money. Your paid search does not sit outside this. Search now expands one query into many and scores your ads on the same trust the AI reads off those sources. Brand traffic has dropped already for most brands; CPC costs have gone up by 15-35%. A brand the model does not trust pays more for the same visibility. The answer and the media plan are one system now, not two.
You cannot understand this on your own
AI search is complex, and it does not reward a quick look. The technology shifts week to week, and the four engines each read the web their own way. You cannot pick it up casually, or in the gaps between other work. It takes real understanding of the technology and the changing landscape. It takes time, effort, focus, continuity and it takes technology of your own, because the surface is too large and too fast to track manually. Treat it as a side task and you set up for failure.
None of this is a project you finish. It is a position you hold. That is the largest change in digital marketing for generative engine optimization. The ownership.
Show More
marketing05-Aug-2026Ambika SharmaPulp Strategy
similar newsMARG ERP ropes in Suniel Shetty as brand ambassador
similar newsNykaa acquires majority stake in Aminu to strengthen premium skincare play
similar newsWalmart completes acquisition of self-service CTV platform Vibe.co
similar newsGreenply launches creative campaign to spotlight anti-termite Plywood
similar newsTed Sarandos meets the creative voices behind India’s next great stories
similar newsTBWA unveils new global identity designed by DXD
similar newsPepsiCo India drops energy branding from Sting amid FSSAI compliance crackdown
similar newsShift to execution stage: We are out of the sandbox stage of AI – Google’s Priya Choudhary
similar newsFSSAI prohibits Dabur from selling food items carrying 100% claims
similar newsUnited Spirits challenges FSSAI labeling prohibition in Bombay High Court
