SEO used to mean reverse-engineering a visible algorithm. Now the mechanism itself is probabilistic, contextual, and personalized down to the person asking.
ESHOW BARCELONA, TECH SHOW MADRID, APRIL 2025
At eShow Barcelona, talking about why SEO is becoming a black box.
Why SEO is shifting from an auditable discipline to one you can only influence.
What happens to visibility when AI systems synthesize the answer before the user ever clicks.
The practices that keep a brand relevant even as the mechanism behind search becomes less transparent.
The same query can return meaningfully different answers depending on someone’s history, location and device. Once there’s a distribution of answers instead of a ranking, optimizing for position #1 stops being a coherent goal.
You can no longer fully audit the mechanism deciding what an AI surfaces, but you can still steer toward being the kind of source it reaches for. That shift from audit to steering is the real story behind the black box.
Just blind.
Victor de la Fuente
For fifteen years my work has focused on where eCommerce strategy meets conversion and emerging channels. Sessions like this one matter because AI is often sold to marketing teams as a shortcut, when it is really a multiplier: it amplifies whatever testing discipline, or lack of it, a team already has.
Speaking and lecturing at top Spain universities, business schools and international events.
I think we’re watching SEO stop being a discipline you can fully audit and start being one you can only steer. For years, SEO meant reverse-engineering an algorithm — find the signal, satisfy it, measure the rank. That model assumed you could see enough of the machine to game it. What’s happening now is different. Search is folding into conversation. A user doesn’t type three keywords into a box anymore; they ask a question, get an answer synthesized from dozens of sources, and maybe never see a ranked list at all. The engine deciding what to surface isn’t running a transparent scoring formula — it’s an AI system weighing context, prior interactions, and personalization signals in ways that aren’t fully visible even to the platforms running them, let alone to the sellers and marketers trying to optimize for them.
That’s the black box I mean: not just “the algorithm is a secret,” which has always been true, but that the mechanism itself is now probabilistic, contextual and personalized down to the individual asking. Two people can ask the same question and get meaningfully different answers, shaped by their history, their location, their device, the conversation they’re already in. You can’t optimize for a single ranking anymore because there isn’t one — there’s a distribution of answers, and your job shifts from “rank #1” to “be the kind of source an AI reaches for when it’s building an answer for someone like this user, in this context.” That’s a harder, fuzzier target, and it’s exactly why the sellers and marketers who treat this shift seriously — rather than waiting for it to resolve back into a familiar keyword game — are the ones who’ll still be visible three years from now.
More speaking events from Victor de la Fuente: Teaching & Speaking
Explore more: Teaching & Speaking · eCommerce & Online Marketing
