Keyword research
Start with customer language
Collect the words customers use in calls, emails, reviews, and support questions. Remove identifying details, then group recurring problems and phrases into a seed list.
The starting point
Before you begin.
Keyword tools can show estimates, but they rarely explain why someone chose a particular phrase. Conversations reveal the problem behind the words, the alternatives a customer considered, and the point where they became unsure.
The process
How to put it into practice.
- Gather repeated questions from sales notes, support requests, reviews, and intake forms. Remove names and private details.
- Copy the customer’s own phrases into a spreadsheet; add columns for the problem, service, and stage of the decision.
- Group similar questions and check which ones your existing pages answer before creating anything new.
Keep a simple research sheet with the exact phrase, the underlying question, the relevant service, and a link to the page that currently answers it. Mark patterns that appear in more than one conversation. Before drafting, check whether the answer belongs on a service page, in a short FAQ, or in a separate guide.
A practical example
What this could look like.
A roofing company repeatedly hears “Can you repair a leaking roof without replacing it?” That question could inform a repair service page or a focused guide.
Review the work.
Keep the evidence behind your decision: the pages you checked, the date range, the data source, and any assumptions. If you make a change, record what changed and check it again using the same method where possible.
Search visibility can move for many reasons, so compare relevant periods and avoid attributing an inquiry or ranking change to one action without supporting evidence.
What to keep in mind.
A frequent question can be useful even when a keyword tool reports little volume; validate it against real customer needs.
Common questions
Frequently asked questions.
What if we do not record sales calls?
Use emails, inquiry forms, anonymized support questions, and public reviews. Ask customer-facing colleagues which questions come up repeatedly.
Should I publish customer wording exactly?
Use it to understand the question, but remove personal details and rewrite examples where needed. Ask permission before using a quotation that identifies someone.
What to do next.
Choose one relevant page or question, document what you find, and agree on the next action before expanding the work. Review the result using the data available to your business.
For the fundamentals, read what SEO is and how it works. If you need hands-on help, explore my SEO services.