eigenKOR is a marketing technology firm, headquartered in Houston with a team of more than thirty people worldwide. Since 2012 we've built customer acquisition systems for 303 businesses — local, regional, national, and international.

Most of what you're hearing about AI and marketing is wrong in a specific way.

The industry has decided this is a content problem. Publish more. Write it for machines. Add an FAQ page. Sprinkle in some schema and call it a strategy.

That is not what decides the answer.

There is a new judge. AI.

Ask ChatGPT or Gemini to recommend a company in your category, and it usually won't answer from memory. It goes looking.

It runs its own searches — several of them, phrased differently from the question you asked — pulls a small set of sources, and builds an answer out of what comes back.

Three things determine whether your company is in that answer.

Can AI find you?

Not for the question your buyer typed. For the questions the model invents on its way to answering. Those are rarely the same, and optimizing for the first while ignoring the second is most of what agencies are currently selling.

Can AI tell who you are?

A machine has to resolve you as a specific entity — distinct from every similarly named company — and attach a set of claims to you with confidence. Inconsistent names, addresses, service descriptions, and categories across your own properties are enough to break that on their own.

Can AI verify what you claim?

A claim that appears only on your website is an assertion. The same claim corroborated across sources the system already trusts is a fact. These systems weight the two very differently, and most businesses have never built the second kind at all.

Content matters at the margin. These three decide the outcome.

Five surfaces, and only one of them is yours

Almost everything written about AI visibility is about websites. That's because the website is the part you own, and the part an agency can bill for cleanly.

It's also about a fifth of what decides the answer.

Those three conditions get satisfied — or missed — across five surfaces.

Your website.

The only one you fully control. Reachable by the crawlers that feed AI answers, meaningful without JavaScript, structured so a machine can parse it, fast enough not to be skipped. Get this wrong and nothing else lands. Get it right and you're at the starting line.

Your identity records.

Every profile, listing, registry, and directory that describes your company — and whether they agree with each other. Cheapest lever available, most often broken. Three versions of the name, two addresses, four different service descriptions, and a machine that can't confidently tell which company it's looking at.

The sources that vouch for you.

Reviews, comparison sites, industry directories, trade press, and the "best in category" roundups these systems lean on heavily. This is where verifiability is actually decided, because a claim you make about yourself and the same claim made by someone else are not weighted the same.

The open conversation.

Communities, forums, Q&A threads, video, social. Heavily retrieved, almost never managed, because none of it belongs to anyone. What gets said about your category in those places shapes the answers about you whether you show up or not.

Your evidence.

Original data, published expertise, earned media — anything specific enough to be worth quoting. This is the difference between being mentioned and being cited, and citation is what returns you as the answer rather than as a footnote.

Which surface matters most is not universal. For a local service business it's identity records and reviews. For SaaS it's comparison and roundup content. For e-commerce it's product data and third-party listings. Anyone who tells you the answer is the same for every business is selling a template.

Being in the answer means being in all five. Not on average. Not eventually. In the answer your customer gets, in the moment they ask for it.

What we work on

Foundation first, every time — the website has to be sound before anything built on it holds. But the engagement doesn't end at the domain boundary, because the answer isn't decided there.

We work the identity layer until every source agrees. We build the corroboration that turns your claims into verifiable facts. We find the sources these systems actually draw from for your category's questions — and get you into the ones that matter. We make your expertise citable.

And we measure the thing itself: a defined set of buying questions, tracked across ChatGPT, Gemini, Perplexity, and Copilot, over time — who gets named, which sources each answer drew from, and where the gap between you and the companies being recommended actually sits.

We report that against your pipeline, not against a dashboard metric only we care about.

It is unglamorous work. It is also where the recommendation is decided.

The automation layer

Being recommended is the front of the system. Keeping the customer, and turning them into someone who recommends you, is where the economics actually live.

We build and deploy automation at three points. Acquisition — so the demand you generate is handled fast and consistently, because a lead that waits is a lead that leaves. Retention — so revenue you've already paid to win doesn't quietly leak. Advocacy — so the customers who are happy say so somewhere it counts.

That third one isn't a soft benefit. Customer advocacy is what fills the third surface: the reviews, mentions, and third-party accounts that turn your claims about yourself into facts a machine will repeat. Advocacy and AI visibility are the same system seen from opposite ends — which is why we don't sell them as separate things.

Who does this work

Marketers who learned how the machines decide.

That's an unusual overlap. Most people who understand these systems technically have never had to sell anything, and most people who can sell have no idea what's happening underneath the answer. The work lives in the middle, and there aren't many of us there.

It also has to be creative work, because this is a moving target. How people research, discover, and choose who to buy from is changing faster than any playbook survives. The questions change. The products change. The sources they draw from change. For most of what we do there is no best practice to copy yet — the first person to work it out is the one who benefits.

The judgment is the part that doesn't automate. Which of the five surfaces matters for your business. Which questions your buyers actually ask, in their words. Which sources are worth the effort and which are noise. That is a human call every time — and it's why this is a practice, not a piece of software.

We also use AI heavily in the work itself — to move faster, to cover more ground, and to read across more sources than any team could manage by hand. That isn't in tension with the paragraph above. Understanding how these systems work is what lets you use them well, and it is the same understanding that lets you influence what they say. One body of expertise, pointed in two directions.

Three responses. We work with one.

Most companies won't act until the pipeline shows it. By then the position belongs to someone else, and taking it back costs a multiple of what holding it would have.

Many will see it and wait for the standards to settle. In most technology shifts that's a defensible instinct. Here it's the wrong call, because the advantage compounds — corroboration accumulates slowly, and whoever starts earlier stays ahead by roughly the amount of time you waited.

A few will move while it's still cheap.

We work with the third group. Not on principle — it's simply where the work returns enough to be worth doing.

The companies we work with

303 of them since 2012 — startups through global organizations, across professional services, healthcare, home services, e-commerce, SaaS, fintech, consumer brands, and companies not yet public about what they're building.

The mechanics are remarkably consistent across all of them. What changes is the competitive set and how buyers phrase what they want — and those two things are most of the strategy.

What we're accountable for

Market share, revenue, profit.

Visibility is the instrument, not the objective. If the work doesn't reach the numbers that determine whether the business is healthy, it hasn't worked. We'd rather be measured against those than against a metric that flatters us.

eigenKOR is a marketing technology firm founded in 2012 and headquartered in Houston, Texas. We help businesses become the companies AI assistants recommend, through technical website optimization and AI visibility engineering, and we deploy automation that improves customer acquisition, retention, and advocacy. Our team of more than thirty people serves clients locally, regionally, nationally, and internationally. We have worked with 303 businesses across professional services, healthcare, home services, e-commerce, SaaS, fintech, and consumer brands. All client engagements are governed by mutual nondisclosure agreements.