The work that brings you customers who are ready to buy
Core GEO is the engagement, and we run it end to end rather than handing you a plan to carry out. YouTube, PR and consulting sit alongside it for the companies that need them. What reaches you is an enquiry from somebody who already knows what they need and what they will spend.
The deciding happens before anyone picks up the phone
Your buyer works the whole thing out in a conversation you never see. The work below puts honest, checkable answers about your business in the places that conversation draws on, so the person who does reach you arrives with the choosing already done.
The main offering, and everything inside it
Five surfaces, run together rather than picked off a menu. What separates one plan from the next is how far we take each one: how many content pieces, how many platforms, how much of the off-site work we run, and how often we review it. The technical work is complete on all of them.
Technical optimization
Every technical fix the site needs, and then the work past fixing: the structure, the markup and the crawler access that decide whether an assistant can read you at all. A site that cannot be parsed cannot be cited, however good the writing on it is.
- Technical audit
- Structured data
- Page speed and Core Web Vitals
- llms.txt and crawler access
- Search and AI readability
Website content optimization
The pages you already have, turned into pages that answer something. Most sites describe themselves and very few answer anything, so we go through what is there and rewrite it into clear, checkable statements an assistant can lift and a buyer can act on.
- Page rewrites
- Answer and entity structure
- Internal linking
- Titles and metadata
On-page content we publish
Every category has a handful of questions people ask when they are close to deciding. We write the clear, honest answer to each one and publish it on your site under your name. Someone who has read those before they contact you turns up knowing what you do, how you work and roughly what it costs.
- Blog articles
- Comparison pages
- Pricing and process pages
- Question and FAQ pages
- Case studies
The platforms you own off your site
Your own accounts on the platforms your buyers already read, written for properly rather than cross-posted into. The same answers, put where people are before they have thought to open anybody's website.
- Medium
- X
The communities your buyers read
Before anyone spends real money they go and see what other people say. We talk about your brand in the communities that matter in your category, meaningfully and under your name, written by people who know the subject. What you get is a buyer who has heard you spoken well of before they ever ask you for a price.
- Reddit threads
- Industry forums
- Q&A communities
- Review sites
We do the work, not just the plan for it
Every surface above is shipped, written and published by us, so the engagement does not turn into a backlog for a team that already has one. Where you would rather hold on to something, you hold on to it, and we work the way your policies say.
We run it, start to finish
Hand over the accesses (CMS, Search Console, Bing Webmaster Tools, GA4, Clarity and the codebase) and the work stops being your problem. We ship the technical changes, write and publish the pages, and post the platform and community work ourselves. Your side of it is an accurate answer when we ask a question and a quick yes on drafts. Nothing here arrives as a deck for your team to implement in a quarter they do not have.
Or on your terms, if access stays in house
Plenty of companies have rules about who touches the codebase or the CMS, and we would rather work inside them than argue about them. Whatever you keep, you keep: your engineers get a specification precise enough to work through line by line, your team gets the drafts ready to publish, and we do not push. One honest caveat: the timeline then moves at your team's pace rather than ours, which is worth planning for.
None of this is an upgrade. It is what makes the work end to end
Eight things come with every engagement whatever it costs: the whole technical program, the rewrites, the audit that starts it, and reporting built on data you own. Some of it scales with the plan (how many pages get rewritten, how far the Reddit work goes), but a plan buys more reach and more published work, not a better version of the floor.
Complete technical work
Schema and structured data, metadata and meta descriptions, internal linking architecture, crawler access and robots directives, page speed and Core Web Vitals, machine-readable files, and anything else on the site that is holding you back.
Website content optimization
Your existing pages rewritten into clear, checkable statements an assistant can lift and a buyer can act on. The plan sets how many: core pages up to fifteen, up to forty, or the full site ongoing.
FAQ content optimization
Rebuilt against the questions your buyers actually ask.
A free AI visibility audit before anything starts
Where you stand across ChatGPT, Gemini, Perplexity, Copilot and Google's AI surfaces. Which answers name you, and which name somebody else.
A dated benchmark report
Your position before we touch anything, written down and timestamped. Everything after is measured against it.
Measurement built on first-party data
Google Search Console, Bing Webmaster Tools and Microsoft Clarity, all on assets you own. Prompt-level sampling covers the engines that publish no first-party data, and we label those readings as directional rather than presenting them as measurement.
Monthly reporting in full
Every source the numbers come from, what moved, what worked, what did not, and what we are changing because of it.
Reddit participation
Under your brand name, with disclosure, in threads where you genuinely have something to add. The plan sets how far it goes: monitored and selective, the full program, or the wider forums and review sites with it.
Two more citation sources, for the companies that need them
Bought on top of Core GEO where the category rewards it. Neither one replaces the work above.
YouTube as a citation source
Video earns its citation twice over: a person watches it, and an assistant reads the transcript underneath. We run the whole thing end to end. We ideate the topics, script them and edit what comes back. The one part we do not do is stand behind the camera, and we support that as hard as we support the rest: shot lists, direction and prompts, so the day you record is a day of reading rather than writing. It is priced per production run rather than by the month.
PR as a source for AI citations
Third-party sources are climbing as a citation source faster than anything else, and that has produced an enormous number of cheap ones. A mention nobody reads does nothing. It is a line on a report, not a citation. We go after third-party mentions that carry real weight in your category, on the straightforward basis that those are the ones an assistant leans on when it answers. No retainer, no upfront and no guarantees. You pay a fixed fee only when a placement actually runs, plus whatever the publisher charges.
AI search, run by your own team
For companies with the people already in house and a preference for keeping the process there. We sit with your team, map out the whole process and the roadmap for AI search, and hand back something they can run.
For your technical team
An in-depth technical report on what the site does now, what is stopping it being read and cited, and what to change. It is written to be worked through line by line rather than admired. Your engineers own the work; the report is what tells them where to start and why it matters in that order.
For your content team
The high-value opportunities in your category, identified and ranked, with actionable steps on how to take each one and a timeline against it. Not a list of keywords: a plan with an order to it, so the team knows what to make next and what it is meant to do once it is made.
Frequently asked questions
How much content do we actually get?
Plans run from 10 to 25 content pieces a month. Each one is published on your own site and then rewritten properly for every other surface that plan covers, so 10 becomes 10 pieces on your site plus 20 rewritten versions across two more. That is three surfaces in all and 30 pieces. At 15 it is 15 plus 60 across four more, which is five surfaces and 75 pieces. At 25 it is 25 plus 175 across seven more, which is eight surfaces and 200 pieces. Not cross-posted. Written for each place.
We rarely recommend going below 10 or above 25. Below that there isn't enough surface area for AI systems to form a consistent picture of you. Above it, the marginal piece starts costing more in quality than it returns in coverage.
The community work is a separate job from any of that. We contribute under your brand name, with disclosure, and a great deal gets discarded before anything is posted. Foundation keeps that selective; Category runs it widest.
The plan sets how many surfaces we publish to and how much of the community work we run. It does not set the technical work: that is complete on every plan.
How do you decide what to publish?
We start by mapping what your buyers actually ask, then check what AI systems currently answer. We identify the high-intent questions in your category, run them across ChatGPT, Gemini, Perplexity, Copilot and Google AI Overviews, and record what comes back: which answers you appear in, which you are absent from, and which companies own the ones you want.
From there we trace where those answers are sourced. AI systems break a single question into sub-queries and pull from a specific set of publications, communities and platforms that differs by category. We map that set for your industry, then build the content plan against the gaps in it.
Nothing gets written because it is a keyword. It gets written because a buyer asks it and the current answer does not include you.
Isn't posting in communities just astroturfing?
No, and the distinction matters to us more than it probably should. We contribute under your brand's name, with disclosure, or we do not contribute at all.
That constraint is why the work takes longer than it looks. We look for threads where your company genuinely has something useful to add, written by someone who understands the subject, and we skip the ones where the only thing to say is a plug. A great many opportunities get discarded on that basis.
The reason is practical as well as principled. Communities remove undisclosed promotion quickly, and a deleted comment is not a citation. Contributions that survive are the ones that were worth posting.
Does this replace our social media team?
No. We publish to a set of external platforms, but only the content that feeds AI search, and only where it is integral to being cited.
We do not run your content calendar, your campaigns, your paid social or your day-to-day community management. Most clients keep their social team exactly as it is, and we work alongside them rather than over them. This is the request we turn down most often, so it is worth being direct about it.
What do you need from us, and what if we want to keep some of it in-house?
We ask for CMS access, Search Console, Bing Webmaster Tools, GA4, Clarity and codebase access, because we execute end to end rather than handing you a document and wishing you luck.
Anything you would rather keep in house, you keep. We will hand your engineers a specification precise enough to work through line by line, and we do not push. Plenty of companies have policies about codebase access and we would rather work within them than argue about them.
One honest caveat. When your team owns the implementation, the timeline moves at their pace rather than ours, and that is worth planning for.
Who actually does the work?
Strategy and direction are centralised, so every account is built on the same method. Execution runs through small teams that never carry more than three clients at once.
That cap is deliberate and it is the main reason we do not scale headcount faster. The work depends on someone knowing your category properly, and that does not survive being spread across a dozen accounts.
How does PR pricing work, and do you guarantee placements?
No guarantees, and no upfront payment either. You pay a fixed fee only when a placement actually runs, plus any costs charged by the publisher.
We built it that way because the alternative produces the wrong incentive. Third-party mentions are climbing as a citation source faster than anything else, which has created a large market in cheap ones, and a mention nobody reads is a line on a report rather than a citation. Charging on delivery keeps us pointed at sources that carry real weight in your category, because those are the ones an assistant leans on when it answers.
PR runs on demand rather than to a fixed monthly quota. We identify the sources worth pursuing in your category, run the outreach and the pipeline, and you pay for what lands.
Who is this not for?
If your buyers do not research digitally, or your addressable market is small enough to reach directly, AI search is not where your money should go.
We turned down a drilling company whose total addressable market was around 150 businesses. High ticket, real budget, and no meaningful case for this work. At that scale direct outreach beats visibility every time, and we told them so. We would rather you spent it on the channel that works than took a retainer from us for a marginal one.