The shift from traditional search to AI-powered discovery is no longer future talk — it’s reshaping marketing right now. Brands that once optimized for Google rankings must now fight for citations and recommendations inside ChatGPT, Perplexity, and other generative engines.
Fresh off research into the LLM marketing landscape, Olayimika Oyebanji sat down with ICODA co-founder and CEO Vlad Pivnev to discuss the real opportunities and hard challenges of Generative Engine Optimization, the limits of measurement in zero-click visibility, and what sustainable AI-first strategies actually look like. Let’s dive in.
Hello Vlad, can you briefly tell the readers about yourself and your route to AI?
My name is Vlad Pivnev and I’ve been in marketing for over 16 years. It started very early — literally selling perfume door-to-door during my first year at university. Later I founded ICODA, a specialized agency focused on crypto and iGaming.
My path into AI was driven purely by data. About two years ago, I started noticing clear shifts in traffic patterns for our clients. People were no longer just Googling — they were asking ChatGPT and getting direct answers. Even more striking, the traffic we saw coming from ChatGPT and Claude converted up to 5 times higher than traditional Google organic traffic. That was the moment I realized this wasn’t something to watch — it was a serious new traffic source we couldn’t ignore.
How would you describe the seismic shift from traditional marketing and SEO to the LLM era? And is this change a fleeting trend, or something more fundamental?
This is not a passing trend — it’s a mass migration in how people discover and choose brands.Google’s desktop search share has fallen to 79%, its lowest level in 20 years. Meanwhile, ChatGPT now handles roughly 17% of all digital queries globally as of Q1 2026. My six-year-old son doesn’t say “Google it” anymore — he says “Daddy, ask ChatGPT.” That tells you everything.
For two decades, we optimized websites for Google’s algorithm. Now we must optimize entire brands for AI models that generate answers. These are completely different systems with different logic. Google ranks pages. LLMs recommend brands. If ChatGPT doesn’t mention you, for most users — you simply don’t exist.I compare this moment to SEO back in 2005. The companies that recognized the shift early and acted fast won big. GEO is at that exact same early stage today.
What are the biggest practical challenges you see companies facing when trying to implement Generative Engine Optimization?
Most companies still have no idea where they actually stand. They assume strong Google rankings automatically mean they’re visible everywhere. That’s simply not true anymore.We run AI Visibility audits across ChatGPT, Claude, Perplexity, Gemini, and others. The results are eye-opening. In the iGaming sector, around 60% of brands are completely invisible to LLMs.
In crypto payment gateways, it’s about 48%. These are companies with solid SEO, real products, and good websites — yet the models don’t know they exist.Then there’s the bigger problem: measurement. There is still no Google Analytics equivalent for LLM visibility. You can’t install a pixel on ChatGPT.
There’s no established playbook either — this is a brand-new discipline that most agencies don’t even offer yet. Waiting is not an option, because right now these models are actively forming their opinions about which brands matter.
Zero-click answers and declining site traffic from AI search are becoming more common. How should brands rethink success metrics and ROI when visibility no longer always translates into website visits?
The entire marketing industry was built around clicks and traffic. That model is breaking fast.Organic click-through rates have dropped by as much as 61% on queries where Google shows AI Overviews. When someone asks ChatGPT or Claude for a recommendation and gets a direct answer, they may never visit your website — yet they’ve already formed a strong opinion about your brand.
This is why brands urgently need to shift from measuring traffic to measuring presence and sentiment in AI responses. Does the model recommend you? What exactly does it say about you? The traffic that does come from AI is smaller in volume, but it converts dramatically better . Fewer visits, but much higher-quality visitors. Your entire ROI model needs to evolve to reflect this new reality.
How big is the gap between traditional SEO performance and visibility in tools like ChatGPT or Perplexity, and what does that mean for marketing budgets?
The gap between traditional SEO and LLM visibility is massive — and surprisingly persistent.You can rank number one on Google and still be completely absent from ChatGPT or Claude.
The correlation between strong Google rankings and LLM visibility is much lower than most people expect.Google ranks pages based on links and technical signals. LLMs form opinions based on how your brand is discussed across the broader internet — expert mentions, authoritative reviews, and consistent citations.
It’s much closer to managing digital reputation than classic SEO.Because of this, companies need to allocate budget specifically for LLM visibility. Even shifting just 10–15% of your digital marketing spend can give you a significant head start, especially since most of your competitors still aren’t doing anything in this space.
Do you expect Generative Engine Optimization (GEO) to remain a specialized discipline, or will it simply become part of standard marketing and SEO practices?
In the short term, GEO will definitely remain a specialized discipline. The tools, methodologies, and skill sets required are quite different from traditional SEO. You need people who truly understand how LLMs process information and how to influence their outputs.In the long term — I’d say within three to five years — GEO will become standard practice, just like SEO did.
In 2005, SEO was a niche specialty. By 2015, you couldn’t run serious marketing without it. The companies that invest early in GEO will enjoy a massive competitive advantage. Those who wait will spend years playing catch-up. That pattern has repeated throughout the history of marketing.
What key lessons from crypto marketing and regulatory shifts have proven most relevant to the AI discovery era?
Two big lessons from crypto have proven extremely relevant.First, crypto taught me how to operate when the rules change constantly. Regulations shift every few months, channels get banned overnight — you learn to diversify aggressively and never depend on a single platform.
That exact mindset is what brands need now as Google’s long dominance begins to crack. Second, and perhaps even more important, is the centrality of trust. In crypto, reputation literally decides whether people will hand over their money. LLMs work in a very similar way — they heavily favor brands with strong, consistent, and authoritative mentions across trusted sources. The discipline of building genuine trust signals that we honed in crypto maps almost perfectly onto effective Generative Engine Optimization.
What is the biggest myth about traditional crypto marketing?
That it's all hype and influencers.
The projects that actually survive do real marketing — performance, content, community, PR. The ones that succeeded long-term were never the ones that just bought a few KOL posts and waited for magic. It doesn't matter what channels you use. What counts is the final result.
Can you share your experience as CEO?
I manage about 50 people across two agencies. The hardest part is not strategy — it's making sure the team grows with the business. When the market shifts this fast, you need people who adapt.
What keeps me going is timing. We're in the first carriages of the AI marketing train. The market is forming right now. The companies that figure out LLM visibility first will define how this works. It's a pretty exciting position to be in if you ask me.
Any parting words?
If your brand is not visible in AI-generated answers today, you're already losing customers. But it's not too late — now is just the start.
My advice is simple. Run an audit. Check how you appear in ChatGPT, Claude, Gemini. See the gap yourself. Once you see the data, the decision becomes obvious.
[story continues]
tags