Compact keywords is Edward Sturm's name for a way of doing SEO around the searches people type when they are ready to buy, hire or get in touch. He teaches the method in his course, Compact Keywords, and most of how I think about pages comes from him. In one episode of his show, he reads a viral Reddit thread on how SEOs choose keywords and says where he agrees. This article covers what I took from the course page and that episode, and how we use it to build pages for local service businesses.
What the course page says compact keywords are
Sturm's course page sums it up in one line: Compact Keywords SEO is "scalable SEO landing pages for high-intent scenarios."
He sets it against blog SEO. Blog SEO answers how, what and why questions for people who are not ready to buy, then hopes to win them over later. In his words, Compact Keywords targets warm searchers and blog SEO targets cold ones.
His other claims on that page:
- The method targets high-intent language most sites ignore. He says that means less competition and easier rankings.
- Because of that, you need fewer words per page and fewer backlinks.
- Warm searchers are ready to act, but most businesses do not target the words they use. He says that means Google cannot recommend those businesses.
- In small niches, he says, "A good deal of Compact Keywords is identifying keywords that are not targeted properly and then simply targeting them properly."
Why so few companies build these pages
A very specific search is typed by fewer people than a broad one. A company looking at a keyword tool sees a small number and moves on. That leaves the search open.
In the episode, Sturm says many high-intent searches have no page made for them, so Google shows the closest match it has. A blog post may rank for a buying search only because nobody built the right page.
He says the best searches are the ones where the person knows what they want but not which business gives it to them. He says these "convert at ridiculous degrees."
And he says a page built for the right search can hold its spot. His own report: SEO tools often called a search hard, he "just targeted it properly," ranked first in one or two weeks and stayed there for years. He showed no data for that on the show. It is his own account, and nobody can promise a ranking.
A bath remodeling example
Take "walk-in shower." That search is big and mixed. The person could be looking at photos, shopping for a kit or pricing a job. Page one is crowded with giant sites. When we looked up "walk in shower cost," page one was cost guide articles from sites like This Old House, not company pages.
Now take "replace tub with walk-in shower" plus a town name. One job, one place, and a person who is ready to hire someone. Few companies build a page for that exact search. So Google shows the closest thing it can find.
A page built for exactly that search, in that town, is a closer match than anything else Google has. That is how compact keywords work for a local business.
How Sturm checks whether a search is open
He does not trust keyword difficulty scores from SEO tools. He looks at Google's results himself. His test, from the episode:
- Search the phrase and look at the top results.
- Count how many put the search words in the page title and the web address. Also check the H1 and the start of the first sentence.
- If few do, the search is easy to win.
He adds two warnings. If page one is all product pages, Google wants a product page, so build that. If a search has been targeted a lot and blog posts still beat the sales pages, people do not want to buy on that search. Pick other searches first.
He also agrees with a commenter that you should not ask an AI how hard a search is unless you give it live search data. Without it, the AI guesses "in a confident sounding way."
One page per search, grouped by meaning
Sturm agrees with another commenter: group searches by meaning, then give each group one page. The commenter's version: one main search per page, plus one or two related searches for context. The rest become FAQs or articles. When a search needs its own page, he puts it in the title, the web address, the H1 and the start of the first sentence.
This is also why a one-page website struggles. When we researched one-page sites, Ahrefs and Search Engine Land both said a single page can rank for only a small group of close searches. When Sturm says single pages rank better, he means one page per keyword, not a one-page site. For service businesses he suggests a services hub with one page per task under it. He also says new pages rank faster when related pages on the same site link to them. Those are his opinions.
How we use this at eezyRank
We build one page for each specific search a buyer types. For a local service business, that means a page for each service, a page for each town the business really serves, and a page for each town and service pair that makes sense. Under each pair sit question pages built from what people really search. Each town's pages sit under that town's page, and visitors can reach them from the menu.
One example of the scale: 1,728 pages for one Florida bath remodeler, each built around one specific search.
We use Sturm's test too, for clients who have our Search Console work turned on. Before we add a new page idea, we check Google's top results and count how many put the search words in their title or web address. Few means the search is open, and those ideas go first. Close variations of the same search are grouped so one page answers them.
For those clients, we read Search Console after launch. When a page shows up for a search but sits below the top spots, we can add a short section built around that search. Searches no page covers become new page ideas.
Not every page will win, and no one controls rankings.
You can see what we build for local service businesses at https://eezyrank.com.
