What does it take to stay visible when search itself is being rewritten?
In a recent feature on LBB, Mohammed Faizan N shares a sharp take on why traditional SEO alone is no longer enough in the age of AI-driven discovery. Find the original article here.
I’ve been having the same conversation with brand leaders for the past six months, and it always starts the same way. They pull up their analytics dashboards, point to a concerning dip in organic traffic, and ask me what’s going wrong with their SEO. The numbers look off. The rankings seem stuck. The traffic isn’t converting as well as it used to.
But here’s the thing: nothing’s actually wrong with SEO alone. The game changed while everyone was staring at the old scoreboard. Search as we knew it has fundamentally shifted, and most brands are still optimising for a world that no longer exists.
The citation problem hiding in plain sight
Let me paint you a picture of what I see across every brand audit we conduct. Recently, I worked with a major wellness and mental health brand that was absolutely dominating its space in traditional search. They ranked in the top positions for their core search terms. Beautiful performance by 2023 standards. Their SEO team was celebrating. The C-suite was happy with the organic traffic numbers.
Then I asked a simple question: What happens when someone asks ChatGPT or Perplexity about wellness and mental health solutions? We ran the test right there in the meeting room. The AI gave comprehensive answers, pulling from multiple sources. The brand wasn’t among them. Not even mentioned. Meanwhile, competitors with significantly worse traditional rankings were getting cited in every single AI answer we tested.
The difference? The competitors showing up in AI answers had a presence across the open web: industry directories, marketplace listings, review platforms, “best of” roundups, etc. The invisible brands, despite ranking well? Just their own websites, beautifully optimised but completely alone. I’ve seen this same pattern play out with clients in food delivery, consumer electronics, and aviation.
That’s the fundamental problem. LLMs no longer trust single sources anymore, no matter how well they rank. They want corroboration, consensus, and repeated mentions across different trusted platforms. We’re calling this the visibility paradox, and it’s playing out across every industry I’m tracking.
21% percent of Google searches now trigger AI summaries, and that number has tripled in just 12 months. But most brands are still optimising for the 79% while completely ignoring the fastest-growing channel in search history. It’s like focusing all your marketing budget on print ads in 2010 while mobile was taking over.
Why SEO isn’t dead but definitely isn’t enough anymore
Every time this topic comes up in a strategy meeting, someone inevitably asks, “So, SEO is dead now, right?” And I get why people jump to that conclusion. It feels like the ground is shifting under our feet. But no. SEO isn’t dead. It’s the foundation. You absolutely need it. It’s just not the whole building anymore.
Here’s how I explain it to clients. AI engines are fundamentally lazy, which works in our favour if you understand the pattern. They strongly prefer citing sources that Google and Bing already trust. Strong SEO gets you in the door. It makes you crawlable, rankable, and findable. But once you’re through that door, completely different rules apply. Traditional SEO gets you discovered. GEO gets you cited and quoted.
And the business stakes here are real. Click-through rates dropped from 15% to 8% when AI summaries appear in search results. Recent studies show clicks down 58% for top-ranking results when AI Overviews trigger. That’s half your traffic potentially vanishing if you’re not present in the AI answer itself.
I saw this firsthand with a global consumer electronics manufacturer. Their traditional search traffic was stable, but brand consideration metrics were declining in key markets. When we audited their AI visibility, the answer became obvious. They were completely absent from AI powered research journeys that their target customers were increasingly using.
The five-pillar framework for AI visibility
We’ve spent the last few months reverse-engineering citation patterns across ChatGPT, Gemini, and Perplexity for over 200 different queries across a dozen industries. From wellness and mental health to food delivery to aviation services. What we discovered is that AI citation success isn’t random. It follows five distinct patterns that, when implemented together as a system, create compounding results.
Pillar 1: Entity stabilisation
LLMs need to confidently identify your brand before they can cite it. Take a company like M+C Saatchi Performance. If it appears as “M&C Saatchi Performance” on the website, “M and C Saatchi Performance” on LinkedIn, “MC Saatchi Performance” in industry directories, and “M+C Saatchi Performance Group” in press releases, the LLM becomes genuinely confused. When LLMs are confused, they simply don’t cite you.
I’ve helped multiple clients fix this exact issue. One airline client had seven variations of its brand name across its digital properties. Once we standardised everything and added proper organisation schema, their AI citation rate jumped 35% within weeks. This isn’t about SEO anymore. It’s about making your brand entity machine-readable and consistent everywhere it appears.
Pillar 2: Answer-cluster content architecture
LLMs automatically break user queries into subquestions. When someone asks about mental health apps, the AI internally asks, “What features matter?” “How much do they cost?” “What do reviews say?” “How effective are they?” Traditional SEO creates separate, thin pages for each variation. That’s backwards now.
We build comprehensive pages that answer the main question plus all related subquestions in one authoritative resource. These answer-cluster pages generate 40 to 60% more citations than traditional keyword pages because they match how LLMs actually process and answer queries. One wellness client restructured just 12 pages this way and saw their citation rate double in eight weeks.
Pillar 3: Strategic secondary source seeding
Here’s where most brands completely miss the opportunity. LLMs are trained to find consensus across multiple trusted sources. One source saying you’re great means almost nothing in their algorithmic worldview. Five trusted sources saying you’re great? Now they’re paying attention.
Working with a food delivery client, we identified that they were present on only their own domain and one marketplace. Meanwhile, competitors were showing up across eight to 10 different trusted sources: industry directories, review platforms, comparison sites, “best of” lists, and community forums. We systematically built their presence across five key secondary sources over three months. Their citation rate doubled. This isn’t link building. This is citation architecture.
Pillar 4: Structure for machine extraction
Content structure has become non-negotiable. If an AI bot can’t easily extract your answer, it won’t cite you. Clear H2/H3 headers that signal topic boundaries. One focused idea per section. Short, direct answers that get to the point immediately. The wandering blog post that takes 800 words of preamble? Dead on arrival.
We layer in structured data: FAQ schema, Product schema, and Article schema. 82.5% of AI citations come from well-structured pages. A global consumer electronics brand restructured its content library with clear sections and proper schema. Within two months, their AI citation rate jumped 45%. Same content, better structure, dramatically different results.
Pillar 5: AI-specific measurement
Google Search Console doesn’t show you AI citations. You need different tools and metrics. We track citation frequency by platform (ChatGPT, Perplexity, Gemini, because each behaves differently). Answer share, which is the percentage of AI answers in your category that mention your brand. Traffic source mix between AI-driven and traditional organic. Visibility scores across LLM engines.
For a wellness client tracking mental health AI adoption, we found that 49% of users were using AI tools in their research journey. That insight reshaped their entire content approach. Within four months, they went from zero AI citations to appearing in 60% of relevant AI answers in their category.
The five pillars aren’t optional tactics. They’re a system. Entity work makes your brand recognisable. Answer clusters make your content citable. Secondary sources provide the consensus LLMs need. Proper structure makes extraction easy. And monitoring lets you see what’s working and double down. Each pillar amplifies the others. Skip one, and the whole framework weakens. Implement all five, and you build a citation moat that competitors struggle to cross.
Why 2026 is the inflection point
The market is moving faster than most marketing leaders realise. AI summary adoption jumped from 6.5% to 21% in just one year. We’re seeing rollout across 200+ countries and 40+ languages. Working with an airline in Southeast Asia, we’re tracking adoption patterns that mirror global trends, just with different platform preferences by region.
The financial stakes are staggering. AI search is projected to influence $750 billion in US revenue by 2028. And 44% of users now prefer AI as their primary research tool, surpassing traditional search engines for the first time.
But here’s what keeps me up at night: early movers are building compounding advantages right now. Every citation builds authority. Every authority mention increases future citation probability. The brands moving aggressively in 2026 are creating a citation moat that gets exponentially harder to cross with each passing month.
The fundamental shift
You’re no longer optimising for rankings. You’re optimising for being citable. Informational content generates six times as many citations as promotional content. LLMs are answering questions, not showing ads.
The traffic from LLM citations is higher quality, higher intent, and further along in the decision journey. Users arriving via AI citations have already consumed your answer, already saw you positioned as an authority, and already have baseline trust. That’s fundamentally different from a cold click from traditional search.
Here’s the existential risk: if you’re not showing up in AI answers, you’re losing consideration entirely. You’re filtered out before users even know to look for you. Most brands don’t realise they’re invisible. They’re watching traditional search rankings while a channel growing 300% year-over-year completely bypasses them.
The brands that adapt in 2026 will own market share in 2027 and 2028. The ones that don’t will spend years trying to understand why their “good SEO” stopped driving growth.
