Building Content That Language Models Quote
This entire area usually amounts to a day of work. It is routinely the difference between a brand that appears in answers and one that does not, and it is worth doing before anybody writes a single word of new content. llm seo
The practical conclusion is unexciting and reliable. Do the work that pays off under multiple scenarios, keep measuring, and treat any strategy that requires one channel's terms to stay fixed as a bet rather than a plan. llm seo
Show the Cheap Failures First Before asking for a programme, ask for permission to check whether you are readable. Crawler access, rendering without JavaScript, listing accuracy on the sources your prompts cited.
Be Honest About What Cannot Be Measured State the limits at the top rather than being caught out on them. There is no console reporting how often you were named. Referral attribution is incomplete because some assistants strip referrer data. Most of the channel's value arrives without a click.
This is the least interesting subject in the discipline and the one that most often explains a total absence from generated answers. A brand can do everything else correctly and remain invisible because a line in a text file, or a setting nobody remembers enabling, is turning the relevant crawlers away.
Keep It Current and Say So Because retrieval happens at answer time, freshness carries real weight. A page updated this month can be cited this month, and a competitor can displace you simply by revising a page you have left alone for two years.
It is also worth recording the reason for every rule you keep. A disallow line with no explanation gets preserved indefinitely through migrations and redesigns because nobody dares remove something they do not understand. A one line comment saying who added it and why turns a permanent mystery into a decision that can be revisited.
This channel is currently less correlated with budget than any other in marketing, and that will not last. The advantages available to a small business today exist because the field is young, the incumbents are slow, and several of the things that matter cannot be bought quickly.
Fragmented identity produces a specific symptom worth recognising: an assistant knows facts about you but attributes them vaguely, or confuses you with a similarly named business. The fix is dull consistency work across every place your name appears.
Ask for a Small, Bounded Commitment Do not ask for a year. Ask for one quarter with a defined scope: run the baseline, fix access problems, correct the listings on the sources that appeared, publish two pages that answer the questions your baseline showed were answered badly.
Stage Two: The Comparison Moves Inside the Machine The current stage is more consequential. A generated answer does not just supply a fact, it performs the comparison the user would previously have done themselves by reading three results and forming a view.
Pick your moment as carefully as your argument. A proposal to investigate a new discovery channel lands very differently in a quarter where organic traffic is soft than in one where everything is comfortable. That is not cynicism, it is recognising that the case is fundamentally about attention, and the same evidence will be received quite differently depending on what else is competing for it.
Publish one honest comparison page naming your real competitors, including where they are the better choice. And start asking every satisfied customer for a review, at the moment they are satisfied rather than a month later.
What llms.txt Proposes It is a proposed convention: a file at your root offering a curated, plain text guide to your site for language model consumers, pointing at the documents you consider authoritative.
A useful way to think about the sequence is that each stage moved a task from the user to the interface. First the fact, then the summary, and now the comparison. Each move removed a reason to visit a website, and each was followed by an industry insisting the change had been overstated. It is reasonable to expect the pattern to continue rather than to stop at a convenient point.
Ahrefs measured this in July 2025 across 15,000 long-tail prompts and four assistants, finding roughly 80 percent of cited pages did not rank for the original query, with about 12 percent in the top ten. The overlap is real but partial, which is the worst case for planning: you cannot ignore your rankings and you cannot rely on them either.
What that implies for planning is modest and unpopular. Any strategy whose success depends on the current interface staying as it is has an unstated assumption in it, and the assumption has been wrong roughly every three years for a decade. Building on the parts that have survived every stage, which are a real product, direct relationships and a reputation independent of any platform, is not a thrilling recommendation and it has an unusually good record.
Include Something Worth Attributing A citation needs something to point at. Passages that contain only sentiment give a model nothing, which is why brand pages full of adjectives are passed over in favour of a competitor's specification table.