How do I tell high-intent keywords from research keywords in a service business?
Sort them into three tiers. Emergency intent uses words like emergency, tonight, near me, no heat, burst. Planned purchase uses replace, install, cost to, contractor, quote. Research uses how to, why is, DIY, symptoms, meaning. Budget and bids should follow the tier rather than the search volume, because volume runs in the opposite direction from intent almost every time.
The three tiers and their tells
- Emergency. emergency, 24 hour, tonight, now, open now, near me, plus problem words that imply a stopped household — no hot water, no heat, burst, flooded, won't drain. These convert fastest and cost the most per click, and they are worth it.
- Planned purchase. replace, replacement, install, installation, new, contractor, company, quote, estimate, best. The customer will buy, but on a timeline of days or weeks, which changes how you measure them.
- Research. how to, why does, what causes, DIY, fix myself, symptoms, lifespan, diagram, error code. Occasionally a research searcher gives up and hires you. Usually they do not.
Cost and price terms are the ambiguous middle
"Water heater replacement cost" looks like research and behaves like planned purchase. Someone pricing a job is shopping, and shoppers buy. "How much does a plumber charge" is closer to research and often price-anxiety.
The distinguishing feature is whether the query names a specific job. Specific job plus cost equals a buyer. Generic trade plus cost equals a browser. This distinction is worth building into your ad copy as much as your bids, because the page that answers the cost question honestly tends to convert the specific-job searcher well.
Volume runs opposite to intent
The research tier has most of the search volume. That is why keyword tools recommend it and why accounts drift toward it over time — the reports look busy and the cost per click is low.
Resist the pull. In a service account with a finite budget, the correct default is to spend the emergency tier out to its limit, fund planned purchase next, and treat research as a content and organic search job rather than a paid one. Research queries are also where answer content earns visibility without paying per click.
Structure should follow the tier
Emergency and planned purchase deserve separate campaigns because they want different budgets, different hours and different bid targets. Inside each, group by service line so the ad can name the job.
The measurement changes by tier too. Emergency keywords should be judged on same-week booked jobs. Planned purchase keywords need a longer window, and judging them on a seven-day conversion window will make them look far worse than they are — a mistake that quietly kills profitable campaigns.
Topics: keyword intent · search intent · keyword research · budget allocation
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.