How do you decide what pages my site needs?
By finding out what people actually type when they want what you sell, and who currently answers them. That research comes before any writing or design, because the pages a business needs are rarely the pages it would have listed from memory.
By looking at what people actually type when they want what you sell, and at who currently gets the work when they type it.
That happens first, before any writing and well before any design. It is the part most builds skip, and skipping it is how a site ends up with beautiful pages about things nobody was looking for.
Why not just ask you
We do ask you. It is the first conversation and it matters. It is also not enough on its own, for a reason that is nobody’s fault.
You know your work in your own words. Your customers search in theirs, and the two are often different. A practice thinks of itself as offering integrative therapy. The person looking types “help with anxiety near me”. A firm lists “private client services”. The search is “who deals with probate”.
Most businesses also underrate the thing they do most often, because it is routine to them. It is not routine to the person searching for it.
What the research actually looks at
What gets searched, and how often. The words, the volumes, and the phrases specific enough to be winnable. Some of what you do has enough demand to justify a page. Some does not. It is cheaper to know that before the page exists.
Who answers those searches now. Usually some mixture of competitors, directories and national sites. It matters which. Beating a directory is a different job from beating the firm down the road. Sometimes the answer is that a phrase is not worth contesting at all.
What those pages actually say. Not to copy them. To find the questions everybody is leaving unanswered, because that gap is usually where the work is.
What you are already found for. If the site exists, Search Console says which pages get shown and for what. There is often a page quietly doing well that nobody realised, and it tells you what to build more of.
What comes out of it
A list of pages, in priority order, each with the job it does and the search it is meant to answer.
Usually that means one page per thing you genuinely do, in the words people use rather than your internal names, plus the pages any visitor needs to trust you: who you are, what it costs, how to start.
It is often fewer pages than people expect. Six or eight pages that each answer something specific will out-earn thirty that overlap. Thirty pages also means thirty pages to keep accurate, forever.
Where the allowance comes in
A build includes six page designs or reusable templates, plus copy for up to ten pages. That sounds tight until you see how it works. A template covers unlimited pages of that type. A services template used for eight services is one design, not eight.
The research is what decides how that allowance gets spent. It is also why the research comes first rather than being a nice extra at the end.
What research cannot do
It cannot create demand. If almost nobody searches for what you sell, the research finds that out. The response is to say so rather than build pages aimed at searches that do not happen. It is a finding worth having early, and it has ended projects here before they started.
Last checked 14 September 2026 · written by Alex Harvey