The glossary is here so that the technical terms that appear in The Revenue Letter have a single, stable definition you can refer back to. Each entry is written in the same plain register as the letters themselves, and is more comprehensive than the in-letter mention. New terms are added as new letters introduce them. If a term is missing, write and tell me which one.

Search engine optimisation

Search engine optimisation, almost always shortened to SEO, is the practice of making a website visible and attractive to search engines. For the last fifteen years it has been the single largest driver of new prospect traffic for most fitness operators. A prospect typed “best gym near me” into Google and scrolled through a page of results. The club that appeared at the top won the click, and the click turned into a tour.

The mechanics of traditional SEO reward longevity. The oldest domains accumulate the deepest backlink profiles. Backlinks act as votes of confidence from other sites, and votes compound over time. Page speed matters. Keyword density matters. Mobile responsiveness matters. The result is a system in which the operators who invested earliest and most heavily in SEO sat at the top, year after year, while newer operators with better products struggled to break through. Age beat quality. Everyone in fitness has lived with this reality.

That model is being displaced. A growing share of prospects now delegate the search to an AI agent rather than scrolling through Google themselves. The agent does not care about your backlink profile. It does not know how long your domain has been registered. It reads your site for what it offers, now, in a format it can extract. SEO is not dead, but the skillset is shifting. The operator who understands structured data and direct answers will be recommended over the operator who spent fifty thousand pounds a year on link building. The competitive advantage that took a decade to build is being unwound in months.

Structured data

Structured data is information about your website formatted so that machines can read it without having to interpret human prose. It lives in the page head as a block of JSON-LD code, invisible to the human visitor but immediately accessible to any automated system that crawls the page. A human sees a logo. An agent sees a structured block that says “this is the Organisation, its legal name is Acme Fitness Ltd, its address is 14 Wharf Road, its telephone number is this, its opening hours are these.”

The common schema types that matter for fitness operators are Organisation, LocalBusiness, SportsActivityLocation, Service, Offer, and FAQPage. Organisation tells the agent who you are. LocalBusiness tells it where you are and when you are open. Service describes what you offer. FAQPage provides direct answers to common questions in a format the agent can cite. Together they give the agent a complete picture of your business without it having to guess what your marketing copy means.

The reason structured data matters now more than it ever did is that AI agents depend on it. A language model can read a paragraph of prose and extract meaning, but it does so with variable accuracy. Structured data eliminates the guesswork. It tells the agent, unambiguously, what you offer, what it costs, and how to contact you. Where structured data is absent, the agent must infer these things from pages that were written for a human reader, and it will often get them wrong or decide they are not present at all. The difference between visible and invisible in an AI-mediated search is frequently one block of JSON-LD in the page head.

Domain authority

Domain authority is a score, originally developed by the SEO tools company Moz, that estimates how likely a website is to rank well in Google search results. It is not a metric Google itself publishes, but it correlates so tightly with actual ranking performance that it has become the industry’s shorthand for a site’s competitive position. The score runs from 1 to 100 and is built primarily on two signals: the age of the domain and the number and quality of other sites linking to it, known as backlinks.

Domain authority compounds. Every year a domain has been live, every backlink it has acquired, every piece of content it has published contributes to a score that becomes harder to shift the higher it gets. A club with a fifteen-year-old domain and a deep backlink profile sat at the top of local search results, and a new independent operator with a better product could not break through because the model rewarded longevity over quality. This was the central frustration of fitness marketing for a generation.

AI agents do not respect domain authority. They do not query a backlink index. They do not know or care how long your domain has been registered. They read the content on your site at the moment of the query and evaluate it for relevance, completeness, and trustworthiness. A two-year-old club with clean structured data and direct answers to prospect questions will be recommended over a thirty-year chain whose website has not been updated since 2019. The metric that dominated discovery for fifteen years is losing its power to a new set of signals, and most operators have not yet adjusted.

robots.txt

Robots.txt is a plain text file that lives at the root of a domain, at the path /robots.txt, and tells automated crawlers which parts of a site they are allowed to access. It is the first thing any well-behaved crawler checks before reading a page. A file that says User-agent: * Disallow: / tells every crawler to stay out entirely. A file that names specific crawlers, like User-agent: GPTBot Disallow: /, blocks only those named agents.

The problem for most fitness operators is that their robots.txt file was configured years ago and never revisited. A developer set it to block all crawlers as a security measure. A WordPress plugin defaulted to aggressive blocking. A marketing agency added a blanket disallow rule during a site migration and forgot to remove it. Nobody has thought about it since, because nobody was losing business over it. That has changed. When a prospect asks an AI agent to recommend a gym and the agent cannot access your site, it cannot recommend you. The prospect never sees your name. The agent moves to the next club.

The fix is straightforward. Check your robots.txt file at yoursite.com/robots.txt. Name the AI crawlers explicitly and grant them access to your publicly visible pages. The major agents to allow are GPTBot (OpenAI), Claude-Web (Anthropic), Google-Extended (Google), and PerplexityBot (Perplexity). This is a text file edit that takes less than a minute. The operators who do it this week will be visible to agents that their competitors’ sites are still blocking.

llms.txt

llms.txt is an emerging standard for giving AI agents a map of a website. It sits at yoursite.com/llms.txt, in the same way robots.txt sits at yoursite.com/robots.txt, and it tells an agent what your site contains, which pages matter, and how to navigate your content. A companion file at yoursite.com/llms-full.txt provides your full factual corpus in a single clean text file the agent can ingest in one pass.

The standard was proposed in 2024 by Jeremy Howard, the co-founder of fast.ai, in response to the growing number of AI agents and language models that read websites on behalf of users. The insight was that websites built for human browsing are inefficient for agents to consume. An agent has to crawl a site page by page, parse HTML, strip out navigation and footer cruft, and piece together what the business actually offers. llms.txt collapses that process into a single file the agent reads first, saving it the crawl and giving it the site’s own description of what matters.

For a fitness operator, llms.txt should contain a brief description of the business, a list of key pages with their URLs and a one-line summary of each, and any other information the business wants an agent to know before it reads further. llms-full.txt should contain the full factual content of the site: facility lists, pricing, opening hours, location details, class schedules, FAQ content, all in one clean Markdown file. Both files together take fifteen minutes to write and deploy. From that point forward, every AI agent that visits your domain will read them, and the quality of their recommendations will reflect the quality of the information you put there.