Frequently asked questions

Good SEO and good GEO are largely the same work. The goal differs: SEO aims to rank a page in a list of links, GEO aims to get your brand cited inside the answer itself. But the inputs are shared. Original research, real topical authority, clean technical architecture, and content that answers what buyers are actually asking all matter to both.

Three things genuinely change. AI systems weigh unlinked brand mentions and entity clarity more heavily than backlinks. They fan out a single question into many sub-queries, so your visibility depends on sources beyond your own site, from industry publications to the communities your buyers read. And they need one consistent version of the facts about your business across every surface, which is why we build a single source of truth before anything else.

What hasn't changed is that thin, keyword-stuffed pages fail in both. Anyone selling GEO as a replacement for SEO is usually selling a tool.

Usually not, though you don't have to decide that on day one. Our scope covers the technical, content and authority work an SEO retainer covers, plus the parts it doesn't: keeping one consistent version of the facts about your business across every surface, and earning placement in the external sources AI systems pull from when they break a question into sub-queries.

Most clients start with their existing vendor still in place. Within the first month they can see the overlap for themselves and consolidate onto one retainer rather than paying twice. Some keep both, and we'll work alongside an incumbent where there's a reason to.

We work directly with heads of growth, heads of marketing, or founders. The opening diagnostic makes the overlap visible either way, so it's a decision you make on evidence rather than one we argue you into.

We baseline your AI visibility before we start, then report against it monthly. Google Search Console now reports impressions inside AI Overviews and AI Mode as a dedicated view, and Bing Webmaster Tools reports Copilot and AI summary performance. We combine those with prompt-level tracking across the engines your buyers actually use, plus on-site behaviour data, so you can see which pages are earning citations and which are not.

The baseline matters more than any single metric. Without a before-picture, month three is just a number with nothing to compare it to.

We'll also be straight about the limits. Google's own AI surfaces are now directly measurable. Traffic from ChatGPT, Perplexity and Claude often lands in analytics as direct, so for those we report citation presence and share of voice against your named competitors rather than claiming a clean click-to-lead trail. Anyone promising exact attribution across every AI surface is describing something the platforms don't currently expose.

Most clients see visible movement within 30 days. The first month is front-loaded with work that changes how AI systems read you: strengthening technical architecture, unifying your information into a single source of truth, closing content gaps, and publishing against the high-intent queries your buyers actually ask. Those are structural fixes, and structural fixes move fastest.

Starting position sets the pace. A brand with an established site and reasonable technical health moves inside a month. One starting from close to nothing takes longer. Our slowest case was a real estate client with no digital footprint at all, not even a Google Business Profile, and that took 90 days to show traction. From there it compounds, because authority and external citations build on each other rather than resetting monthly.

Speed also depends on you. We need accurate information, access to help us extract original data, fast approvals, and clear communication. Clients who turn drafts around in days rather than weeks see results proportionally sooner.

Plans are $999, $1,500 and $5,000 a month. Each starts with a 90-day commitment, then runs month to month. The difference is reach, not quality: every plan gets the same technical work, the same writing standard and measurement on data you own. What changes is how much of your site we cover, how many pages we publish (ten, fifteen or thirty a month), and how far beyond your website we go. The pricing page sets out exactly what each plan includes. Above the three sits a custom plan, for groups running several brands, markets or legal entities at once.

Before any of that, the 90-day roadmap is free to anyone who asks for it, client or not. It is not conditional on a call going well and there is nothing attached to it.

We ask for ninety days because that is a fair test. The technical foundation lands in the first month and the content compounds after that, so one month tells you very little. After ninety days, you continue month to month or you stop.

We work with brands where the buyer researches before they commit: D2C and ecommerce, B2B from technology through manufacturing, SaaS, and property developers. Those are the categories where someone asks an AI to compare options before they ever reach a website, which is where the work pays off hardest. We take clients outside these verticals when the buying pattern fits.

The method is the same across all four. What changes is which queries matter, which external sources AI systems pull from in that category, and who you're being compared against.

Opening publishes 10 pages a month on your site, Middlegame 15 and Checkmate 30. A page is either new, or a new version that replaces an outdated one, and each month's pages are chosen from the questions your buyers ask where you are missing. We never write a page just to reach a number.

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. Community work is part of Checkmate.

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.

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.

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.

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.

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.

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.

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.