The problem
You know customers are searching for what you do — you just don't know the exact words they use, which of those words you could realistically win, or which pages to build. Most "keyword research" you can buy doesn't answer that either: it's a spreadsheet export from a volume tool with zero judgment applied. Ten thousand rows, no plan.
What we do
We do keyword research the way you'd want a skeptical engineer to do it. For every candidate phrase, we look at the actual search results page — who currently owns it, whether they're entrenched ten-year specialists or beatable generic content, and whether the searcher is ready to buy or just browsing. Then we score each phrase for winnability: can your site, with its history, realistically earn page one — and in what timeframe?
The output isn't a spreadsheet, it's a battle plan: which phrases to own, which to feed with supporting articles, which to skip because chasing them would burn a year. And we're honest about a thing most vendors aren't — when we don't have legitimate volume data for a phrase, we say so instead of inventing a number. Qualitative judgment labeled as judgment beats fake precision labeled as data.
How it works
- Seed harvest. Your services, your service area, your customers' actual language (we mine reviews, calls, and competitor pages — not just tools).
- Live SERP analysis. For each cluster, we read who actually ranks today and how strong they are.
- Intent and winnability scoring. Every phrase rated: buyer-ready or research-mode, winnable or entrenched, head term or long-tail opening.
- The page map. Each winning phrase mapped to a specific page or article to build — with no two pages competing for the same phrase.
What you get
- A scored keyword map: phrase, intent, competition read, winnability, and the reasoning
- A page-by-page build plan (pillar pages, supporting articles, internal-link structure)
- An explicit "do not chase" list — the phrases that would waste your money, and why
- Everything explained in plain English, ready to hand to any writer or developer
Receipts
For a California mortgage brokerage, we researched and validated 689 individual city rows — loan-limit data, local context, search behavior — feeding a local-page engine now publishing in waves. For our own site, we built the same map we'd sell you: live SERP-verified across 20+ queries and 8 topic clusters, with zero invented volume numbers anywhere in the document. We plan our own money pages off this method.