Forward-deployed, not handed off
We work inside your operation rather than at arm’s length from it — one person doing the work himself, a systematic method, and AI doing the volume work so judgement goes where it actually matters.
Most marketing work fails in the gap between the recommendation and the change. A consultancy delivers a deck. An agency delivers a report. Somebody at your end is then supposed to translate it into edits on a website they may not control, alongside their actual job. Six months later nothing has shipped and everyone is disappointed.
We work the other way. We deploy into the problem, do the work where the work is, and stay until it is live and measured.
Four things that make this different
Embedded, not at arm’s length
We work alongside your marketing lead, your web person and whoever owns the product data. Not a monthly call and a PDF — actual collaboration on actual pages, in whatever tools you already use.
One person, start to finish
Lance runs your tests, diagnoses the cause, writes the fix and reports the result. There is no junior to hand the work to after the pitch, because that hand-off is where accountability disappears.
Systematic, not intuitive
The same sequence every time: test, diagnose, fix, re-test against the same fixed questions. Every claim traces back to a transcript or a URL. When something does not move, we say so rather than reframing it.
AI does the volume; people do the judgement
Running hundreds of queries across engines, crawling sites as a machine sees them, comparing months of results — that is automated, and it is why this is affordable at all. Deciding what your buyers actually ask, and what a result means commercially, is not.
What a month actually looks like
No mystery, and no discovery phase that bills for six weeks before anything happens.
| Stage | What happens |
|---|---|
| Test | Your fixed set of buyer questions run through ChatGPT, Perplexity and Gemini, with the model versions recorded. You get transcripts, not a summary. |
| Diagnose | Where you were absent, we work backwards from what got cited instead. The competitor’s cited page is a better specification than any general advice. |
| Fix | Technical remediation, structured data and content, written and implemented. Delivered as changes to your site, not as recommendations for someone else to make. |
| Re-test | Same questions, same engines, same models. Progress is a number that moved or did not. |
We are not a software development shop, and we do not present ourselves as one. We do not write applications, produce CAD or BIM assets, or run paid media. Engagements are scoped to work we can execute well ourselves — which is why the deliverables list is shorter than most agencies’ and why the things on it actually get done.
Why this is possible now
Ten years ago this work would have needed a team: analysts running queries, a technical SEO auditing crawls, writers producing content, someone building the reporting. The economics only worked at agency scale, with the layers of account management that agency scale requires.
That has changed. The repetitive parts — the querying, the crawling, the comparison, the first draft — are now genuinely automatable, and we have built our process around that rather than bolting it on. What is left is the part that was always the valuable bit: knowing which questions matter, reading what the evidence means, and deciding what to do about it.
It means one experienced operator, with AI doing the volume, can deliver what used to take a department. That is the whole reason a one-person firm can do this work at this standard.
We published a full case study on a 26,000-product catalogue we own — the diagnosis, the measurements either side of it, the AI visibility test, and the defects still outstanding. Read the Western Sport case study ›
Before you ask
Are you an agency or a consultancy?
Neither in the usual sense. An agency sells you a team you rarely meet; a consultancy sells you a recommendation and leaves. We work forward-deployed — embedded with your people, doing the work alongside them, and staying until the change is live and measured.
How big is the team?
One person. Lance Rankin does the work himself — the diagnosis, the fixes and the reporting — and uses AI for the volume work that used to need a team. There is no hand-off after the pitch, because the hand-off is where most of the value leaks out.
You say you use AI heavily. Does that mean the work is automated?
The volume work is: running hundreds of test queries, crawling sites, comparing responses across engines and months. The judgement is not. Deciding which questions your buyers actually ask, and what a result means for your business, is the part that determines whether any of it is useful — and that is done by a person.
What do you not do?
We do not write software, build applications, produce CAD or BIM assets, or run paid media. We scope engagements to work we can execute well ourselves. Where something falls outside that, we say so rather than subcontracting it and marking it up.
What do you need from us?
Someone who can publish changes to the site, and a subject-matter expert available for questions about specifications and terminology. We handle the diagnosis, the structure and the writing; you confirm the technical facts are right. Typically a few hours of your team's time per month.
Find out what AI says about you
We run the buyer questions in your category through ChatGPT, Perplexity and Google Gemini, and send you a short report showing whether you appear, who appears instead, and why. No cost, no obligation, no sales call required to receive it.
Request your free checkUsually back within two business days.