A large share of searches now end without a click, answered by an AI summary at the top of the page. For a services company this is not the end of search as a channel; it is a change in what wins. The pages that get cited in the summary — and get the clicks that remain — are the ones that answer a specific question completely, show real expertise, and are structured so a machine can find the answer.

Write for the question, not the keyword

Each page targets one question a buyer actually asks — 'how long does a CRM migration take', 'what does a voice agent cost to run' — and answers it in the first paragraph, then supports the answer. Keyword lists are a starting point for finding questions, not a template for stuffing. A page that answers ten questions badly ranks for none.

Show that a person with experience wrote it

Specific examples, numbers where you have them, what went wrong, what you would do differently. Named authors with real roles. Consistent, honest opinion. This is what search engines mean by experience and expertise, and it is also what makes a buyer trust the page.

Structure for machines

Clear headings that are the questions. Article, FAQ, Service, Organization and Breadcrumb structured data. Fast pages, mobile first, no interstitials. An llms.txt describing the company. Sitemaps that include every article. These do not make bad content rank; they make good content legible.

Measure citations and conversions

Rankings still matter but citations in AI answers and branded search volume are the new signals of authority. Track which pages are cited, which queries produce branded follow-up searches, and which articles precede enquiries. Publish on a cadence you can sustain — a good article a week beats ten thin ones a month.