The Disambiguation Competition You’re Already Losing
Strategy Sandbox - jasonbarnard.com
The problem isn’t that Google and AI platforms can’t find you. For most people and many brands, they can find you. The problem is that when they do, they’re not sure which “you” they’ve found.
That is a completely different problem. And it has a completely different solution.
Google’s job isn’t to find you - it’s to decide which entity you are
Google filed a patent that describes the “apple” problem. The word “apple” maps to at least two concepts: a fruit and a computer company. When someone searches for “apple,” the system doesn’t just retrieve pages about apples. It determines the intended meaning, assigns the query to the most relevant concept identifier, and builds the response accordingly.
That process is called disambiguation, and Google has been engineering it for twenty years. US7925610B2 describes how related concepts and contextual signals guide the meaning determination. US10528871B1 extends it: a token like “Apple” relates to multiple concepts, and the system uses similarity and pertinence signals to assign it to the dominant one.
The machine’s fundamental job isn’t retrieval. It’s interpretation. And interpretation begins with disambiguation.
Most people and many brands are unresolved entities in that system
An unresolved entity is one where the machine hasn’t confidently assigned the dominant interpretation. It knows the name. It may have found dozens of pages about you. But it hasn’t determined, with confidence, that you are the authoritative version of that name across its systems.
The machine hedges. “Claims to be.” “According to their website.” “May refer to.” These aren’t style choices. They’re linguistic signals that disambiguation hasn’t completed.
You aren’t invisible. You’re ambiguous. That’s a harder problem, and most people have no idea it exists.
People experience ambiguity as obscurity - and misdiagnose the cure
When I talk to accomplished professionals about how Google and AI describe them, the complaints are consistent. “It says wrong things.” “It confuses me with someone else.” “The AI says ‘a consultant who claims to specialise in X’ rather than just stating it.”
Every one of those is a disambiguation failure. But the person describing it doesn’t reach for that word. They say “I’m not visible enough” and go looking for more content, more press, more SEO. They’re treating obscurity when the diagnosis is ambiguity. The two require entirely different interventions.
For companies with common or descriptive names, the problem looks clearer on the surface. “Horizon Financial” competes with Horizon the TV show, the airline, the insurance group, and at least three SaaS startups for the same interpretive real estate. The ambiguity is obvious, the competition is visible, and the brief practically writes itself.
For a person, the competition is invisible. But it runs every time their name is queried.
Trademark law solved this 100 years ago - under a different name
Trademark law has long recognised the doctrine of acquired distinctiveness: descriptive or otherwise weak identifiers can, through sustained market association, come to signify one source in the minds of the public. When “Corn Flakes” or “Holiday Inn” achieved that status, courts recognised the association as protectable.
Algorithmic Acquired Distinction is the machine-layer analogue: the point at which search engines and AI systems resolve an ambiguous identifier to one dominant entity. The concept has a strong parallel in trademark doctrine, even if the mechanism is entirely different. One hundred years of legal precedent says this kind of battle is winnable.
For me, the clarifying moment was recognising that the same sustained, corroborated, distinctive association that earns trademark protection in human minds is exactly what earns dominant entity status in machine systems. Different arena. Same underlying logic.
Algorithmic Acquired Distinction is the strategic objective - disambiguation is just the mechanism
The full chain runs like this.
Ambiguity is the enemy. Every person and brand with a non-unique name starts from a position of interpretive competition. The machine must pick a dominant interpretation. If you haven’t earned that position through deliberate, consistent, corroborated signals, someone else has, or nobody has and the machine hedges indefinitely.
Disambiguation is the machine’s process for resolving ambiguity. It’s not something you do. It’s something that happens to you. Your job is to influence the inputs it uses.
Algorithmic Acquired Distinction is the strategic outcome. When the machine has resolved your identity with high confidence, consistently selects you as the dominant interpretation, and states your attributes as facts rather than hedging them, you’ve achieved acquired distinction.
From that foundation, everything else follows. Reliable understandability depends on successful resolution. Credibility depends on stable understandability. Deliverability depends on both. The chain holds, but only if disambiguation runs first.
Every AI visibility tool measures the scoreboard - nobody is training the players
The current market has thirty-plus platforms tracking citations, monitoring mentions, scoring share of voice, measuring how often AI platforms include you in responses. These are useful. They tell you where you stand in the disambiguation competition.
None of them train the inputs that determine how that competition resolves.
That’s the structural difference. Monitoring tools observe outputs. Algorithmic Acquired Distinction engineers inputs. One tells you the score. The other changes the outcome.
The machine must pick one. Make sure it picks you.
Jason Barnard is the founder of Kalicube® and coined the terms Brand SERP (2012), Answer Engine Optimisation (2017), and Algorithmic Acquired Distinction (2025). Kalicube builds Algorithmic Acquired Distinction for people and brands.