Module 3: Keyword Research

Keyword research is where strategy becomes a list you can execute: the terms worth chasing, judged by intent and value rather than raw volume, and mapped — one page, one job — to the site. This module covers where good terms hide, how to judge them, and how to build the map every later module works from.

Finding terms: where good keywords hide

Start with a seed list: the plain-language phrases a customer would use for what the business sells — not industry jargon, not brand names. "Emergency plumber Dallas" is a seed. "Residential repiping solutions" is a brochure. Write twenty to thirty seeds from the business's services, locations, and customer problems, then expand each one with four reliable wells of real search language.

Competitor pages are the first well: the titles and headings of the pages already ranking in your market are a free list of terms Google takes seriously, and any tool or manual SERP review will show you gaps — services they cover that your client doesn't. "People also ask" and related searches at the bottom of the results page are the second: real follow-up questions in searchers' own words, which make excellent page sections and standalone guides. Google's autocomplete is the third: type a seed slowly and note the completions, which reflect actual search demand.

The fourth well is the richest and most beginners skip it: Search Console itself. For any site that's been live a while, the Performance report lists the queries the site already gets impressions for — including terms ranking on page two that nobody deliberately targeted. A page sitting at position 11 for a valuable term is usually a cheaper win than a brand-new page chasing a term from zero. Mine the client's existing data before you research anyone else's; the gold is often already in the account, waiting for one honest refresh.

You're doing keyword research for an established bakery site. Where do you look first for the cheapest wins?

Judging a keyword: intent, difficulty, value

A long candidate list isn't a strategy. Every term gets judged on three questions, in this order. Intent: what job is the searcher doing, and can this business's site honestly do that job with a page? A term with the wrong intent is worthless at any volume — this is the Module 2 test, applied one keyword at a time. Difficulty: who ranks now, and how strong are they? If page one is all national brands and huge directories, a local business won't break in this year; if it's other local businesses with thin pages, it might. Be honest about the timeline — your credibility is spent every time you chase a term you knew was out of reach.

Business value is the question beginners forget. A search that ends in a $5,000 job is worth more than fifty searches that end in a free download. Estimate it crudely but deliberately: what does a conversion on this term pay, and how likely is the searcher to become one? "Cost to replace a water heater" signals a homeowner mid-decision with a budget; "water heater facts" signals a student. Volume matters — a valuable term nobody searches earns nothing — but it ranks third, not first.

The practical output is a short priority list, not a spreadsheet of thousands. Ten well-judged terms with clear intent, beatable competition, and real value will occupy a small business's SEO for months and produce results you can report. A bloated list produces scattered work and a client asking why nothing moved. Judge hard now so you can ship confidently later.

For a roofing company, which term deserves priority? Term A: "roofing" — enormous national volume, ranked by Wikipedia and national brands. Term B: "roof replacement cost" in their service area — modest volume, ranked by other local roofers with thin pages.

Mapping: one page, one job

The keyword map is the deliverable this whole module builds toward: a simple document — spreadsheet or doc, the format doesn't matter — where every priority term is assigned to exactly one page, with the intent noted beside it. One page, one job. The rule sounds obvious and is violated constantly: two service pages chasing the same term split the site's signals and often rank worse than either would alone, and one page stuffed with five unrelated terms does none of them well enough to rank.

Build the map in three passes. First, list the site's existing pages and assign each the term it should own — most pages get one, a few get a primary term plus a close variation. Second, mark the gaps: priority terms with no page to own them become the content plan, ordered by the value judgments from the last lesson. Third, mark the collisions: terms two pages both claim. Collisions get resolved by merging pages, retargeting one page to a different term, or differentiating the intent — a "cost" page and a "repair vs replace" page can coexist; two identical "services" pages can't.

The map then runs the engagement. On-page work in Module 4 edits pages to fulfill their assignment; content briefs in Module 7 are written straight off the gap list; and your monthly report can honestly show terms moving against the plan you made. Do the worksheet below on a real sample business — your practice site's niche is perfect — because the first map you build will be slow, and it should be slow on a site that can't fire you.

Your turn: Build a starter keyword map for a sample business (your practice site's niche works). List 8–10 priority terms. For each, write: the term, its intent (informational / navigational / commercial / transactional), the business value in one line, and the page that owns it — existing page, new page to create, or "collision" if two pages would claim it. Flag one page-two term from Search Console if your lab site has data already.

A client's site has two pages — "Heating Services" and "Our Heating Services" — both targeting "heating repair" terms, and both stuck on page three. What's the standard fix?

What you can now do

Build a seed list from a business's services, locations, and customer problems, then expand it into the language customers actually search with.
Judge any keyword on intent, difficulty, and business value — in that order — and cut the terms that fail the intent test at any volume.
Hand over a keyword map where every priority term is assigned to exactly one page with one job, so pages stop competing with each other.