AI Search Optimization: Why Internal Buy-In Is Harder Than the Technical Work
Summarize this article with AI
103 words on ChatGPT. 4 words on Google. The real divide is organizational
103 words.
Four.
I’m not making this up. Crystal Carter, Head of AI Search at Wix, cited these numbers at SMX Advanced. The average ChatGPT prompt is 103 words. On Google, a query runs 3 to 4 words. Two worlds. Two logics. Two levels of complexity.
You’ve felt it in meetings. On one side, the technical SEO team digging into schemas, entities, response formats. On the other, marketing that still sees AI as a funny ChatGPT for churning out emails. And in the middle, leadership. The people who approve budgets. The ones asking: « what’s this actually for? »
I watch this gap every week with my clients. From my office in Southeast Asia, I look at sites that could capture AI traffic. They have the pages. They have the authority. But nothing moves. Why? Because no one internally built the coalition that turns an AI SEO project into a business priority.
The technical work is clear. Structuring content, tagging sources, producing pages that agents can parse. It plans. It quantifies. The bottleneck is elsewhere. It’s sitting around the table. It’s called internal buy-in. A game of numbers and narrative, not code.
Face this friction head on. Once leadership is aligned, everything else flows downhill.
My e-commerce client had the technical team ready. The CMO blocked everything
Here’s a concrete case. A client, 9,800 products. Technical catalog, lots of product sheets, solid SEO already. The technical team called me one Tuesday morning. They wanted to prep for AI Overviews and conversational agents. They’d already mapped the pages, schemas, entity structure. A full action plan ready in 6 weeks. Budget: €22,000.
I arrived on the project. I validated feasibility. I said: « I show you how to build this, you deploy it. » Everything rolled. Except the CMO hit the brakes. He didn’t see the return. He said: « AI is moving fast, we’ll look next year. » For him, it was a buzzword. For the team, an opportunity slipping away.
I spent four weeks building a presentation with data, not promises. I dug into competitors, pulled AI Overviews already triggered in their sector, showed that smaller brands were already getting cited in those responses. I calculated that across 47 informational queries tied to their products, 31 triggered an AI Overview. Zero mentions of their brand. Zero.
I showed that conversational prompts on ChatGPT already mentioned their products, but responses pointed toward comparison sites, never their product sheets. We weren’t losing non-existent traffic; we were forfeiting traffic about to be born.
The CMO flipped when I said: « You don’t want to be cited when someone types 103 words describing exactly your product? » The following week, the project was approved. Result after 7 months: 4,700 organic sessions per month from AI. Sessions no one had predicted the year before.
The technique? Flawless. Internal buy-in? The real fight.
Memory vs. personalization: the Carter framework and why it shakes teams
At SMX Advanced, Crystal Carter drew a sharp distinction. On one side, inferred memory. What the AI assistant deduces from your exchanges: tone, complaints, habits. On the other, declared personalization: preferences, linked accounts, authorized apps. Memory shapes style. Personalization changes what the agent accomplishes.
Why does this shake teams? Because you can’t optimize for a single AI response. You have to target signals that cross both layers. It’s not metadata business. Your brand needs presence across connected ecosystems, linked Google accounts, structured profiles. That touches product, CRM, marketing automation. So teams that don’t do SEO.
What struck me was the experiment run by iPullRank and cited by Carter. Three accounts. Same prompts. Different levels of personal data. Result: AI Mode responses visually distinct, with specific recommendations. One account got a streaming suggestion addressed to a child by first name because the tool knew the family. It’s controlled, documented, concrete.
Want to convince leadership? Show them that experiment. Say: « This isn’t theory. AI is already personalizing responses. If our brand isn’t in the signals, we don’t exist for those users. » Now you’re not talking SEO. You’re talking customer expérience. Buy-in shifts registers.
Jen Cornwell at SMX Advanced: « Internal coalition is the only lever that accelerates an AI roadmap »
Right after Carter, Jen Cornwell, Senior Director of AI SEO at Tinuiti, hammered home the point. She wasn’t talking schemas or predictions. She was talking internal coalition. Based on her expérience across dozens of brands, an AI roadmap moves when three conditions align: an executive sponsor, a point person per department, and a shared dashboard of AI signals.
Translation: SEO alone has zero chance of securing budget for a cross-functional AI project. You need the content manager, brand lead, IT, CRM. That’s the only way the signals you send to AI agents stay consistent everywhere.
Practical checklist I see work at my successful clients:
- A CMO who grasps that AI is a layer added to search, support, purchase—not just another channel.
- An IT leader who opens APIs to connect product sheets to AI surfaces.
- A brand manager who reformats content for entities, not humans.
When those three are aligned, technical work deploys in weeks. Without them, it stays in limbo.
Leadership speaks the language of numbers, not SEO
Let’s get concrete. How do you get buy-in? Not through SEO. Through commerce, finance, marketing. Show opportunity in their language.
First number: 103 vs. 4. A user types 103 words. That’s specific intent. They want a product, solution, recommendation. If your brand isn’t in the response, you lose a high-intent sale. That decision traffic is no longer free.
Second signal: the percentage of queries with AI Overview in your sector. For the e-commerce client, I found 31 out of 47. That’s 66% of their informational queries already triggered an AI result. A year earlier, it was 0%. Leadership doesn’t move for « AI is the future. » They move for 66%.
Third angle: competitive advantage. I showed that a smaller direct competitor was starting to appear in AI Overview citations. Why? Structured buying guides as entities. No ad budget, no link wars. Just semantic architecture built for AI. That makes a CMO nervous.
Build a mini-dashboard of these three signals:
- Evolution of AI Overviews across your 30 key queries
- Your brand positioning in conversational responses (ChatGPT, Gemini)
- Competitive benchmark on AI citations
Vous avez besoin de parler le langage des chiffres ? Voici le plan en 10 slides que Jen Cornwell recommande pour aligner les équipes et décrocher l’approbation. Chaque slide correspond à une étape clé du pitch.
Les 10 slides qui convainquent votre direction en 10 minutes
Déroulez la présentation impactante proposée par Jen Cornwell pour obtenir le budget AI
How to make AI matter on the internal agenda without wasting time
Another Cornwell insight I’ve rolled out with all my clients since January: the « 10-10 presentation. » 10 slides, 10 minutes. No jargon. No crawling. Just:
- Slide 1: what AI does in our sector (concrete AI Overview examples)
- Slide 2: our current presence (zero or close)
- Slide 3: the conversational prompt gap (103 words)
- Slide 4: monthly opportunity cost (lost sessions × estimated average order value)
- Slide 5: three immediate actions (schemas, entity content, brand signals)
- Slide 6: team needed (sponsor, IT, content, SEO)
- Slide 7: 90-day roadmap
- Slide 8: investment and first measurable milestone
- Slide 9: risks of doing nothing
- Slide 10: the question: « Who owns this project for us? »
In 10 minutes, the message lands. It’s an enterprise project, not an SEO ask. The e-commerce client I mentioned adopted this format. Leadership said yes by slide 9. No need to convince for six months.
The beauty of it? This format forces SEO out of its technical corner. They become a business project lead. AI agents demand brands orchestrating their signals, not siloed teams.
You don’t need to wait for AI to stabilize
One last blocker I hear: « AI changes too fast, let’s wait for it to stabilize. » Wrong. AI won’t stabilize in the way you mean. It’ll personalize more and more. It’ll infer more and more memory. Which means the longer you wait, the more your competitors capture the edge of declared signals.
Experiments like iPullRank’s show that personalization gaps are already here. Not prepping for them means letting others define which brand becomes the reference in AI responses. It’s an enterprise decision, not a technical problem.
So ask this Monday morning. Who in your organization grasps the 103 vs. 4 differential? Who knows that 66% of your strategic queries show an AI Overview without citing you? If the answer is nobody, the hard part isn’t code. It’s the conversation you need to start now.
Live audit of your AI presence in 60 minutes
I don’t sell you the method. I show you the pages. For an hour, I project your sector: AI Overviews, chatbots, missing signals, architecture to build. You leave with a roadmap to convince leadership. No jargon.
Book a strategic call — 45 minFrequently Asked Questions
Why is internal buy-in more important than technique for AI optimization?
AI signals touch brand, product, and CRM. Without an executive sponsor and cross-functional collaboration, the technical SEO can’t deploy schemas, entity content, and the connections needed to appear in personalized agent responses.
How do I convince my leadership that AI Search isn’t just hype?
I see this constantly with clients: the percentage of your key queries already triggering an AI Overview, the traffic you’re forfeiting, the intent gap between a 103-word prompt and a 4-word Google query. A 10-10 presentation with no jargon is enough to flip decisions.
What’s the distinction between memory and advanced personalization that Crystal Carter described?
Inferred memory (tone, habits) drives response style. Declared personalization (linked accounts, preferences) changes content. Brands must work both—it goes beyond SEO alone.
Which departments should be involved in an AI Search project?
In practice: SEO, content, brand, IT, CRM at minimum. Jen Cornwell recommends an executive sponsor, a point person per department, and a shared AI signal dashboard so everyone stays aligned.
Where do I start on AI optimization without big budget?
I recommend mapping AI Overviews across your 30 strategic queries. Identify pages that could be cited, structure them as entities, and strengthen your brand signals on your Google Business profile. Low cost, high impact—and it builds the data to persuade internally.
