Six roles, and what each one delivers in week one.
Almost every brand we measure already has a marketing team. What it does not have is someone who knows why the assistant names the competitor: of the 45 Portuguese B2B SaaS companies we audited, 96% had no declared policy for AI crawlers at all. These six roles join the team that already exists, for six weeks or for six months, and bring the measurement with them: eleven engines every week, in the same dashboard we look at.
Six roles, and what each one does in week one.
Week one is written down on purpose. It is what separates hiring a skill from hiring a promise: after five days something is done, and it is the same something for everyone who joins in that role.
GEO lead
Designs the plan across the eight dimensions and decides the order: what gets fixed first, what gets published, who gets pitched. Owns the number the brand wants to move.
Week 1: Reads the last few weeks of measurement, names the three questions where the brand loses, and writes the plan by horizon.
Technical SEO
The foundation GEO runs on: HTML that reads without JavaScript, correct schema.org, llms.txt and robots for AI crawlers, Core Web Vitals, indexing.
Week 1: Runs the site scan, separates what stops an engine reading the page from what merely slows it down, and fixes the first group.
Technical content editor
Writes the pieces engines cite: honest comparisons, original data, direct answers to the question the buyer actually asks. Sober register, no adjectives standing in for numbers.
Week 1: Takes the question with the worst citation rate and writes the piece that answers it, with the source in plain sight.
Digital PR and entity
Everything off-site that decides whether the brand exists to an engine: press, podcasts, directories, Wikidata, name consistency everywhere.
Week 1: Maps where the brand is named today, where it should be, and opens the first outreach list.
Measurement analyst
Connects Search Console, GA4 and Bing, separates traffic arriving from AI assistants, and answers the hard question: did this produce revenue?
Week 1: Connects the missing integrations and writes the reading of what was already there and nobody had read.
GTM and revenue operations
Carries visibility into the funnel: which questions are worth money, how a mention becomes a conversation, what sales does with a lead that arrived through an AI answer.
Week 1: Maps the measured questions by decision stage and names the ones that touch revenue.
Three shapes. Duration is what changes.
Sprint
Six weeks, one named deliverable
A closed problem, one of our people solving it and leaving the method behind. It ends with a delivery and with the client's team able to repeat it.
Fits when: There is an identified gap and nobody internal with time to close it.
Part time
Two days a week, three months minimum
The person joins the team's routine: meetings, reviews, plan. Not consulting seen from outside, work done from within, with the rest of the week free for other clients.
Fits when: The work is continuous but does not yet justify a full-time hire.
Dedicated
Full time, six months minimum
The person works for one brand only, with our team behind them: whatever they do not know, someone here does, and that is not billed separately.
Fits when: GEO has become a permanent front and the brand is not ready to hire.
The conditions, before you ask for them.
Thirty days' notice once the minimum is served
The three and six month minimums exist because below them the work does not have time to pay off. Once served, thirty days ends it, with no penalty.
We pick who goes in, and we can say no
The conversation starts with the problem, not with CVs. If we have nobody who fits that problem at that time, we say so rather than sending whoever is free.
Measurement is included and never billed separately
Whoever rents one of our people gets Periscopy access on day one. That is where you see what they changed, week by week, without depending on the report they write themselves.
Each model closes on a named deliverable
The sprint ends with a delivery and with the method written down for the team to repeat. Logged hours with nothing to show at the end are the most expensive way to buy time.
Four variables set the price.
A rate card per profile would give everyone the wrong number: the same editor costs one thing for six weeks and another for six months, and a GEO lead who also answers for the client’s team is not the same product as a pair of hands. Four things set the price, and we name them before proposing: the profile, the seniority, the duration, and whether there is exclusivity.
What does not vary is Periscopy access, included in all three models.
30 minutes to know whether it makes sense.
Tell us what you want solved and who you already have on it. We leave that call with three answers: which model fits, which profile goes in, and what they deliver in week one. If the brand does not need anyone of ours, we say so and it ends there.