Your buyers ask ChatGPT, Claude, Gemini and Perplexity before they ask Google. When the answer names a competitor you lose the deal with no auction, no ranking and no line in any report. Nobody in your company owns that today, and your agency does not measure it. This is not a monthly report and a call. It is hours a day inside your team, heaviest in the first months while the fleet gets built, and the fleet keeps the work running 24/7, because the difference between an engine that recommends you and one that never names you is decided in hours nobody is watching.
No call required to see your numbers. Three concurrent seats, and I say no when they are full.
AI visibility is not a campaign you buy once. It is a standing position that decays every time an engine changes, which is monthly. It needs an owner with authority inside the company, not a supplier with a scope.
Citation share is not a ranking. The tools your team already pays for were not built to measure which brand an engine names in an answer, so it never appears in the monthly deck.
Entity graphs, schema clusters, llms.txt, agent readiness. It is engineering work with a marketing outcome, and the people who have shipped it are not on the job boards.
Engines change, competitors publish, your own content shifts. A one-time audit is a photograph. This needs somebody watching the position and reacting to it.
Hours a day from me, heaviest in the first months, because that is what building the plumbing and the fleet actually takes. And the work itself is always: buyers ask engines at midnight, competitors ship on weekends, answers change overnight. That is why we automate, so the seat runs 24/7 even when nobody is at a desk. Daily presence in your Slack, anchored by a standing weekly session your team can plan around. The only thing on a monthly rhythm is the read-out; anything else run monthly finds out about a problem three weeks after it cost you something.
Not a status call. Time on your calendar where the work happens with your team in the room: what moved, what we are changing, and who is doing it before the next session.
A fixed prompt matrix across four engines, scored against three named competitors, running day and night rather than when someone remembers. Same questions every week, so the trend means something.
Not office hours. Questions get answered the same day, and a competitor move on a Tuesday does not wait for a Thursday call.
Specific pages, specific questions, structured so an engine can extract an answer. Your writers keep writing; they stop guessing at the format and stop waiting on a monthly deck.
Schema, sameAs cluster, knowledge-panel and Wikidata status, llms.txt and agent readiness. The plumbing that decides whether an engine can resolve who you are at all, kept current rather than shipped once.
When an engine changes how it cites, you get the read within days and the adjustment inside the same month, instead of finding out next quarter from a traffic dip nobody can explain.
One page of numbers and one page of what to do. If there is a board or a CEO above you, it is written so they can read it without translation. If the top seat is yours, the same artifact becomes the training material for the team beneath you. Either way, the weekly work is mine to run; this is the proof of it.
Rankings, structured data, crawl health, and the surfaces people check before they buy. The method that makes an engine cite you is the same method that ranks you, run at a finer grain than a standard SEO retainer, so the old channel is covered without a second vendor.
I join your marketing meeting, take questions from your team and argue with your agency when the answer is wrong. That is the part a vendor cannot do, and it is why this is a seat and not a scope.
Visibility work produces a standing pile of things a person should not be doing by hand: running the same prompts weekly, watching competitors, drafting to a structure, checking whether a page still resolves. Those become agents, built on your stack and scoped to what you actually need. And it does not stop at marketing: anywhere the business has leverage to give, from sales follow-up to finance ops, gets the same treatment. It is part of the retainer, not a separate engagement. The work is always, which is why we automate: the fleet gives the company 24/7 capability, and my hours compound instead of repeating.
Every seat above is a real template in the studio. The button loads the live product into this page: click around, open a template, watch a dry run. What your company gets is this, scoped to your functions and wired into your systems.
An executive assistant agent for a single person, wired into the tools they already live in. Inbox triage, briefing docs before meetings, follow-ups that actually go out. One useful agent beats a diagram of twelve.
Research, sales, support, finance and reporting agents mapped to your real org chart, with the handoffs between them defined, up to a chief of staff agent that runs the sub-agents and reports only to you. If a function has repeatable judgment in it, it can carry an agent.
Connected to the tools your team already uses rather than a new platform to log into. Your systems, your accounts, your keys. Access is scoped to what each agent needs and pooled nowhere else.
Review gates, evals and a defined failure behavior, so output gets checked before it reaches a customer. An agent nobody trusts gets switched off in week three, which is worse than not building it.
The weekly prompt matrix on this page is not me sitting there typing questions into four engines. It is an agent, it is yours, and it keeps running after the engagement ends.
Some companies want the fleet handled end to end; some want to be walked through every step until it is theirs. Both are the job, whichever you prefer. Every agent ships with a runbook either way, so you are buying capability that stays, not a dependency that leaves when I do.
This is the part that does not appear on any other advisor's page. I did not read about entity engineering, I ran it on myself first, and what I built is a live surface that engines already resolve. An advisory client is inside that surface, not outside it.
All of it is distribution and structure that already works, pointed at your brand.
There is still no starter package, no diagnostic tier and no lite version. A cheaper option mostly gets sold to people who were going to buy the real thing. The choice below is not quality, it is reach: how much of the company the agents cover. Both seats get the same weekly cadence, the same depth and the same person.
The marketing seat
The complete engagement, scoped to marketing: everything on this page, all traditional SEO and reputation management, plus AI integrated across the whole marketing function.
The fleet seat
Everything in the marketing seat, plus a fleet of agents across the business, inside and outside marketing, running 24/7 and maintained by me for the term.
Both seats: three-month minimum, thirty days' notice after it, and part of the same three concurrent seats total.
One founding seat is open on different terms, in exchange for a named case study and a reference call. One, and it is not repeated.
Every month should leave your team able to do more of this without me. If I have not made myself less necessary by month three, do not renew.
Each seat gets hours of my day, every day, and I am one person. Three is also the number that keeps this the work I love: building great things with great people. When the third seat is taken the answer is no, not a junior.
Methodology and tooling are licensed for use on your account, never transferred. Your data, your content and your accounts remain yours entirely.
The prompt matrix, the scores and the trend live in your systems and update weekly. Decisions get made on numbers your whole team can open, not on a vendor's summary of them.
Most people selling AI visibility run some prompts and guess. I engineered the real thing for a company I own. Every line below is checkable in about a minute, which is the only kind of credential that belongs on a page arguing for verifiable structure.
No testimonials on this page on purpose. I have not run enough of these engagements to have real ones, and inventing them on a page about verification would be the one mistake this argument cannot survive.
Work first, paperwork second. You see your own numbers before you talk to me about money.
Your brand, scored. Send your domain and two competitors. Twenty buyer questions across four engines, back inside 48 hours, free. If the engines already name you, you have saved a call.
Twenty minutes on the phone. We read the result together. I tell you whether this is worth an advisor or whether it is a two-week fix your team can run alone.
The seat, or nothing. $5,000 a month for the marketing scope, $7,500 with the full agent fleet, three-month minimum either way. There is no smaller step to sell you, so this call is a yes or a no and both are fine.
Month one is the heavy one. Daily hours: the tracking agent goes live and runs itself from then on, the entity-graph gaps get named, the first working agents land in your team's hands, and you have a 90-day plan. If month three has not made your team more capable without me, do not renew.
Your domain and two competitors. Twenty questions across four engines, back in 48 hours, free and with no call attached.
✓ Got it. Your scorecard lands within 48 hours.
Three concurrent seats. No call required to get the scorecard.