“Today external conversations have become readable by algorithms. The impact is bigger, but the principle isn’t new. Your brand is what the ecosystem says about you: it has always been that way,” Lily told us
Here we are again on SEO Confidential, the column where we talk with the most authoritative international SEO experts.
We created this space to understand how search is changing in the age of AI, LLMs and AI-powered answer engines. Together, episode after episode, we look at the concrete repercussions this shift has on your brand and on the visibility of your business.
Today, search is no longer a plain list of links. It is a living entity that answers, synthesizes and mediates, shaped by ever more sophisticated Artificial Intelligence models and LLMs
They are the ones deciding what will emerge, what will be cited and what, instead, will be silently ignored.
The click, once sacred, is no longer guaranteed. Traffic, once idolized, is no longer the center of the universe. Zero-click is not an exception but the new normal. In this unstoppable scenario, your brand transcends your website: it becomes the sum of what the digital ecosystem whispers about you, of what AI models are able to recognize as authoritative.
It is precisely in this context of profound transformation that we had the privilege of talking with Lily Grozeva, an independent SEO and AI search consultant with over twenty years of experience.
Lily Grozeva approached AI search as a strategic discipline in its own right, well before its impact was universally recognized. Her vision, forged by decades of experience with B2B technology companies, offers a unique and invaluable perspective.
Lily is a pioneer in this field, which is why we are truly delighted with this interview. And we are sure you will find plenty of insights to help you find your bearings and, why not, even change your mind about many things you take for granted…
With these necessary premises out of the way, all that’s left for me is to wish you happy reading!
“Direct relationships are the only distribution channel you truly own. You need to prioritize an offline, personal approach to building visibility”
In recent years many companies have chased the idea that AI could solve the resource problem by mass-producing content. All things considered, how much of that promise was marketing and how much was genuinely scalable without sacrificing quality, internal approvals and brand consistency?
Mostly marketing. I have been working on AI search optimization since 2021, before “AEO” was even a widespread term. Back then, with the agency I worked for, we even produced a case study with Jasper, which was the most popular GPT wrapper at the time. The pattern was always the same: AI scales volume, not judgment.
Even back then we understood that the hardest part of content has never been the writing. The real crux lies in deciding what to say, what to leave out and how to align it all with marketing goals. These are choices a machine cannot make: you need a person. And the moment approvals, brand or subject-matter-expert review and legal checks come into play, the “infinite content machine” turns into a very expensive way to produce material that neither the company nor the client truly feels comfortable publishing.
And that’s without even getting into the SEO implications of large-scale content production.
Many teams say they “want to use AI”, but when it comes to execution, blocks, fears and resistance kick in. Is today’s real bottleneck technological or cultural — that is, inside organizations and their dynamics of power and accountability?
It is neither a purely cultural issue nor an exclusively technological one: it is both, but there is a third factor that often gets overlooked — organizational structure.
The actual work of producing content — briefs, SEO recommendations, post-publication optimization, topic validation, business alignment — takes a competent content operations person a couple of hours per piece. What takes weeks, and derails entire projects, is the absence of engaged stakeholders, messy approval chains, and the lack of a subject matter expert available when needed.
That is the real bottleneck. It doesn’t matter how aligned the content team is or how top-notch the automation: you still have to deal with organizational chaos, unclear responsibilities and understaffed teams. AI solves none of this.
If clicks and traffic are becoming increasingly misleading indicators, how does a CMO hold their ground when the board asks for “old-style” numbers? Which KPIs would you use to prove the brand is winning visibility and trust even when search happens without a click?
For me, you have to stop defending traffic and completely reframe the conversation. When I work with B2B clients today, the metrics we focus on are branded search trends, the frequency and quality of citations in AI-generated answers, share of voice in the third-party sources buyers actually use to inform themselves, and pipeline opportunity attribution from discovery calls. These are not “exotic” metrics: they are simply honest ones.
Traffic is a vanity metric. Citations are a sanity metric. If an AI model spontaneously recommends you, unprompted, that is what KPI dashboards should start measuring. The problem is that, for now, they don’t know how yet.
Today discovery no longer happens only through your website: what the ecosystem says about a brand (communities, platforms, discussions, third-party sources) matters more and more. Isn’t this a structural loss of control for B2B marketing, which until now has built its strategies on “proprietary” messaging and landing pages?
It was always an illusion. Chris Walker was talking about the “dark web” probably five years ago already. In B2B, nobody ever bought because they loved your hero section: they bought because three people in their network spoke well of you, because a case study felt credible, or because an analyst mentioned your name in a briefing.
That has always been the real buying signal. It just wasn’t easy to measure, so we measured clicks and pretended they were the same thing.
What has changed today is that external conversations have become readable by algorithms. The impact is bigger, but the principle isn’t new. Your brand is what the ecosystem says about you: it has always been that way. Twenty years of SEO prove that backlinks mattered from the very beginning. Now they also matter because they end up in the models’ training data.
If the click really is dead and search is becoming an AI-mediated decision layer, aren’t we asking marketers to accept a structural loss of control, going from owners of traffic to mere extras inside summaries generated by third parties?
Life as an extra is not mandatory. If you produce commodity content, you are an extra, period.
The real question is whether the organization is willing to do the harder work: building an authentic point of view, proprietary data, hands-on customer experience, a perspective that can only come from having actually done things. A colleague recently recommended Ikujiro Nonaka’s book, The Wise Company: How Companies Create Continuous Innovation, precisely because the approach it describes is exactly what would allow an organization to thrive today.
Personally, I chose to run a small, deliberately lean consultancy, because in this new context depth beats volume. From a founder’s perspective, I would rather be the single sentence an AI cites as authoritative than the ten pages it politely ignores.
You say mentions are replacing backlinks and that authority is built off-site. Let me play devil’s advocate for a moment: doesn’t this risk penalizing smaller brands above all, forced to compete on platforms and media they don’t control, where access is often paid or gatekept?
It’s actually the opposite, if you play the game well. Big brands have resources, but they also have vague, abstract offerings, committees, rigid brand guidelines and approval cycles that make it almost impossible to produce true originality quickly.
Small brands that manage to be the only authentic, competent voice on a niche problem have a real structural advantage in a world driven by AI-based search.
You can’t compete with Salesforce on scale, but you don’t need to. It’s enough to out-think them in a specific niche and be faster. It’s not even a new concept. Search has been a mature market for a long time, and new or smaller players have to approach it differently than they did ten or twenty years ago.
If content marketing becomes “answer engineering”, with content designed to be extracted and reused by AI, isn’t there a risk that brands end up working for free for systems that monetize the answer without giving back either traffic or relationships?
Honestly, it’s not that different from what SEO has been for the past twenty-plus years. We have always built content and offerings to please Google enough to gain visibility, knowing full well that that content was being monetized through advertising. AI search platforms are doing the same thing, including the ad formats arriving as early as this year.
We have long been feeding ranking machines with our content while they monetize it. The difference, until now, was that users still landed on our sites and we had more control over the message. There was no synthesis layer reinterpreting the brand on our behalf.
As brands, and as the professionals who support them, we are still in a position to build offerings, messaging and visibility capable of reaching the buyer, wherever they are. The only real difference is that it’s harder today. But, once again, search is a mature market and growth inevitably becomes more complex.
The split between traditional SEO and AI visibility seems to dramatically increase complexity, skill requirements and costs. Aren’t we effectively telling companies that the search of the future will only be sustainable for those with budgets, teams and maturity far above average?
AI visibility and traditional SEO are part of a single organic visibility strategy, not two separate disciplines. The foundation is still classic SEO. Realistically, 60–80% of the work consists of traditional SEO activities that must be executed rigorously. AI visibility is built on top of this framework.
You can think of it as an updated or evolved SEO strategy, but the substance doesn’t change: it’s still the same strategy. The traditional fundamentals aren’t replaced — they are leveraged and extended to also ensure visibility in AI systems.
Like every evolution, it increases complexity, requires new skills and brings additional costs. The key point, though, is the framework: one integrated strategy — not a migration, not a reinvention.
For many organizations, the keyword was the last measurable, defensible handhold. How do you govern a strategy based on conversations and prompts without sliding into total arbitrariness?
Instead of asking “are we ranking for these keywords?”, the question becomes: are we the go-to voice on this specific set of problems for this specific buyer?
It is not an arbitrary question — if anything, it is more demanding, because it forces you to be honest about what you truly know and what you merely hope to be associated with.
If every piece of content you produce can be traced back to a limited number of non-negotiable truths about the business and its real competencies, then you have a defensible strategy. If that link isn’t there, it’s not a strategy — it’s just an editorial calendar.
If within a few years visibility is decided by proprietary AI systems rather than by users choosing what to click, the risk isn’t just for marketing but for the market as a whole: are we heading toward a future where a handful of models decide who exists and who doesn’t, with no transparency and no real chance to compete?
This is the risk our industry is overlooking — or at least it isn’t tackling the issue with open, honest debate.
If three models influence enterprise purchasing decisions and those models rely on opaque training data and commercial incentives, then market visibility becomes structurally dependent on a black box. That is a competition problem, not just a marketing one.
The most concrete answer any B2B brand can give today is to diversify aggressively. If your customers only know you exist because a prompt surfaces you, you don’t have a brand — you have a listing.
In this context, direct relationships are not optional. They are the only distribution channel you truly own. If your gut feeling is that a market like this pushes you to prioritize an offline, personal approach to building visibility, that instinct is worth trusting.
In 2026 only what is truly proprietary, patented or tied to non-replicable human expertise will survive absorption by LLMs, so… Aren’t we effectively admitting that a large part of B2B marketing (and of the products it promotes) risks becoming irrelevant because it is too easy to summarize, copy or replace?
Yes, and it should be said without mincing words. Much of what has passed for B2B content strategy has always been replaceable. AI simply made it free.
If a GPT can summarize your white paper in three bullet points without losing anything meaningful, then it wasn’t a white paper — it was a long email.
What endures is what always had real value: original research based on hard data, points of view built on implementation experience actually lived through, and that specific human friction that comes from having made decisions in real conditions, taking responsibility for the outcomes.
Decades of field experience cannot be obtained with a prompt. That is the only competitive advantage that truly holds.
From “vanity metrics” to “sanity metrics”
Stop chasing clicks, because that race is already over and you risk being left behind.
The truth that emerged from our conversation with Lily Grozeva is as brutal as it is necessary: Artificial Intelligence isn’t simply updating marketing, it is demolishing the old foundations to rebuild them from scratch. If you still measure your success in traffic and search volumes, know that you are clutching mere vanity metrics while the world has changed currency.
Yes, because the new currency is the citation.
In a future where algorithms no longer just index pages but answer users directly, your brand is no longer what you shout in your ads, but what the ecosystem whispers about you.
Being cited means being recognized as an authoritative source, a reference point the machine considers worthy of trust. And this is not a loss of control, but your greatest opportunity for differentiation.
While AI can flood the web with endless, scalable content, it will never be able to replicate your critical judgment, your firsthand experience or the uniqueness of your thinking.
The secret to not being swept away by algorithmic homogenization lies right here: in being frighteningly human and unrepeatable.
You must have the courage to invest in what no software can simulate, such as your proprietary data and the authentic relationships built outside the usual playbook.
The age of machines throws down a challenge: you must be braver, more original and deeper than you have ever been before. Only what has a soul and a unique point of view will manage to stand out from the crowd and truly be heard.
The question is no longer how to rank on a search engine, but how to become so indispensable that the intelligence itself chooses and prefers you.
See you next week for a new episode of SEO Confidential!
#avantitutta
