What is Keyword Research?

Keyword research: discovering what people actually search

Keyword research is the discovery and evaluation phase of SEO—not the plan that follows, and not the label you paste into a CMS field. It answers: What do people type? How often? With what intent? Can we realistically win?

Research produces raw intelligence: query lists, SERP snapshots, difficulty estimates, and intent tags. Keyword strategy consumes that intelligence to decide what gets built. Seed keywords are inputs to research, not outputs. Confusing these stages leads to briefs with pretty spreadsheets and no publishable priorities.

What keyword research produces

Research vs strategy vs brief roles

A practical keyword research workflow

Evaluation criteria beyond search volume

High volume alone is not a research conclusion—it is one input.

Keyword research methods by source

SERP-first research starts with a live Google search: Who ranks? What formats appear? What questions sit in PAA? This grounds tool data in reality.

Analytics-first research starts with Search Console and on-site search logs. You already earn impressions; research explains whether to double down or fix mismatch.

Competitor-first research maps rival URL structures and heading patterns—useful for gap finding, dangerous if copied without differentiation.

Stakeholder-first research interviews sales, support, and product. Tools miss jargon customers actually use.

Blend all four; no single source is sufficient.

Clustering queries during research

Example cluster for "crawl budget":

crawl budget
what is crawl budget
crawl budget optimization
how to reduce crawl budget waste
crawl budget vs crawl rate

Research decides: one informational hub, one technical guide, or merged page—based on SERP overlap and intent, not word count.

Common keyword research mistakes

• Exporting 10,000 keywords without intent tags or deduplication
• Treating tool difficulty scores as gospel without manual SERP review
• Skipping zero-volume queries that match high-intent niche buyers
• Researching keywords without a business lens—traffic that never converts
• Ending research at discovery without handoff to strategy and briefs

Research deliverables that strategy teams need

• Cluster name and representative query
• Intent tag and recommended page type
• Volume range and difficulty notes (with SERP screenshot)
• Existing URL to optimize OR flag for new content
• Related clusters for internal linking
• Excluded queries with reason (wrong intent, out of scope)

Keyword research and the tail spectrum

Research surfaces queries across the demand curve—from head terms and short-tail keywords to mid-tail and long-tail keyword variants. Research's job is to catalog the spectrum; strategy's job is to choose where to compete given resources.

Do not conflate "long-tail" with "easy." Some long queries sit on authoritative SERPs because intent is valuable.

Refresh cadence

Re-run focused research when:

• Major product or pricing changes
• SERP layout shifts (new AI overviews, feature dominance)
• A competitor publishes a definitive guide in your cluster
• GSC shows rising impressions on queries you did not brief

Stale research produces briefs targeting last year's SERP.

Keyword research myths

• Myth: "Research means picking the highest volume keyword." Reality: fit, intent, and winnability matter more.
• Myth: "Once researched, keywords are fixed forever." Reality: query language evolves; refresh.
• Myth: "More keywords in the spreadsheet equals better research." Reality: prioritized clusters beat raw dumps.

How Crawlox helps keyword research validation

Crawlox connects crawl and index data to content reality: URLs that research assigned to a cluster but never went live, thin pages targeting researched terms without substantive sections, and duplicate templates competing for the same query set. Research is only as good as the site architecture it lands on—Crawlox shows where discovery and execution diverged.

The practical takeaway

Keyword research is the process of finding and evaluating search queries—intent, demand, difficulty, and fit—before anyone writes a brief or picks a primary keyword. Hand results to keyword strategy for resourcing decisions; keep research distinct from CMS labels like focus keyword fields.