Choosing seed keywords when you don't know a client's industry
A fast, repeatable way to find the handful of seed keywords that actually produce a useful research session for a client's industry you've never worked in.
A lot of freelance SEO work starts the same way: a new client signs, and you're expected to start finding keywords for an industry you've never worked in — plumbing, veterinary compounding, B2B logistics software — after maybe an hour of onboarding call. You don't need to become an expert in the field. You need a fast, repeatable way to find the handful of seed keywords that will actually produce a useful research session.
Where seed keywords actually come from
Don't start by guessing terms. Start by extracting language that's already been said out loud.
The client's own words. Whatever the client called their business or service on the onboarding call — often the version they used before they knew any "SEO terms" — is usually closer to how a real searcher phrases it than anything you'd invent.
Their homepage and service pages. What they call their own offerings tells you the client's mental model of their own business, which matters later — you don't want to recommend ranking for terms the client won't recognize as theirs.
Their competitors' service pages. Ask directly who they consider their competition, then pull two or three of those pages. Competitor language is often closer to market-standard phrasing than a smaller or newer client's own site.
One real customer question. Ask the client what a customer asks right before they buy. One good answer is worth more than an hour of generic brainstorming.
From seeds to a research session
Once you have three to five candidate seeds, resist researching all of them broadly and sorting later — that's the fastest way to burn a research session on volume instead of relevance. Instead: run each seed individually, look for real volume and difficulty your client's site can realistically compete on, and save winners directly into the client's project as you go.
Difficulty is worth a second look here specifically because it's a relative signal, not an absolute one: a 42 in one tool isn't a 42 in another, and it says nothing about whether this client's site can realistically compete — that still means opening the top five results and reading them.
Stop once you have a working set for the first tracking cycle. You can always research another seed later; narrowing scope this fast protects your time and your research budget for the sessions that matter.