Keyword research for buyer intent, not just volume

Sort every keyword by intent first

Search intent comes in four types, and the type matters more than the volume. Informational queries ("how does SEO work") seek answers. Navigational queries ("WordStream login") look for a known site. Commercial queries ("best CRM for small business") compare options before buying. Transactional queries ("home addition contractor Pasadena," "bathroom remodeler near me") signal a homeowner ready to act. Informational searches dominate raw volume, but they sit furthest from a signed job. Before you rank a single keyword by search count, tag it by intent. A 200-search transactional term is worth more than a 20,000-search informational one.

Intent is readable from the words and from the SERP. Modifiers like "best," "top," "vs," and "review" flag commercial intent; "buy," "price," "near me," and "for sale" flag transactional. The results page confirms it: if Google shows shopping ads and product pages, it has decided the query is transactional, and an informational blog post will not rank there. Local intent is bigger than most teams assume; roughly 46% of Google searches carry it, and 76% of people who run a "near me" search visit a business within a day.

Map keywords to the funnel, not just the page

Every keyword belongs to a stage of the buyer's journey, and each stage needs a different page. Top-of-funnel awareness terms are informational and broad. Middle-of-funnel consideration terms are commercial: comparisons, alternatives, "best of" lists. Bottom-of-funnel decision terms are transactional: pricing, demos, "near me," branded queries. The mistake is pointing one page at all three. A pricing page will never rank for "what is marketing automation," and a beginner's guide will never convert someone searching "HubSpot vs Salesforce pricing." Match the page format to the stage the keyword sits in.

This also explains why so much content underperforms. Gartner found B2B buyers spend just 17% of their total purchase journey meeting with potential vendors, doing the rest as independent research across digital channels. If your keyword map is all top-of-funnel, you capture attention during that research but never the moment of decision. The fix is balance: enough informational content to be found early, and deliberate commercial and transactional pages to be chosen late.

A workable split most teams can build toward:

Long-tail is where the buyers are

Head terms are short, high-volume, and brutally competitive; long-tail terms are longer, more specific, and lower-volume individually. Backlinko's analysis of 306 million keywords found 91.8% of all queries are long-tail, with a median keyword drawing just 10 searches a month. Added together those long-tail terms account for only 3.3% of total search volume, which is exactly why most teams ignore them. That is the opportunity. A homeowner typing "cost to finish a basement with a wet bar" has already decided what they want; one typing "basement" has not.

Specificity is a proxy for intent. The more words someone adds, the further along the buying cycle they usually are, and the easier the term is to rank for. You will never beat the national lead aggregators for "remodeling," but "primary suite addition contractor" plus your town is winnable and pulls a homeowner closer to booking an estimate. WellBuilt builds keyword maps that target dozens of these precise long-tail terms rather than a handful of vanity head terms, because the aggregate traffic converts and the individual pages actually rank.

Read keyword difficulty before you commit

Keyword difficulty estimates how hard it is to rank on page one, scored 0 to 100 by tools like Ahrefs and Semrush. Ahrefs bases its score primarily on the backlink profiles of the current top 10, so it doubles as a rough effort estimate: a keyword scoring KD 40 in Ahrefs suggests you will need links from roughly 56 different referring domains to compete. Semrush uses a wider formula that also weighs domain authority and SERP features, which is why the same keyword often scores higher there. Treat both numbers as directional, not gospel.

Pair difficulty with intent before deciding. A low-difficulty, high-intent long-tail term is the best work you can do; a high-difficulty informational head term is usually the worst. Note that Google's free Keyword Planner only shows volume in broad ranges unless you run an active Ads campaign, so most teams pull difficulty and exact volume from a dedicated tool. The goal is a shortlist of terms you can realistically rank for that also sit near a purchase, not a spreadsheet of impressive volumes you will never win.

Prioritize the terms that convert

Commercial and transactional keywords are the ones worth fighting for, and they are also the ones holding up best against AI search. Zero-click searches climbed to roughly 68% of US Google by early 2026 as AI Overviews answered more questions in place, but that pressure falls hardest on informational queries. Commercial and transactional intent still sends clicks, because someone comparing vendors or ready to buy wants to reach a page, not read a summary. Weighting your roadmap toward those stages protects you from the traffic that is quietly disappearing.

Build the prioritization into the math. Rank candidate keywords by a simple blend of intent, achievable difficulty, and business value rather than volume alone. The #1 organic position still earns about a 27.6% average click-through rate, so winning a single high-intent term outperforms a fistful of mid-page informational rankings. Capture the commercial and transactional demand first, then layer in the informational content that feeds your clusters and builds the topical authority those money pages need to rank.