Keyword research is not about finding the highest-volume term and writing a page for it. It is about mapping real questions to real content, at the right depth, for the right audience. After 14 years doing this for clients across dozens of industries, that framing is the single most important shift I can offer you. Most businesses I work with come in fixated on a handful of vanity terms. We spend the first few weeks reorienting around what their actual buyers are searching, which is almost always different from what leadership assumes.

Start With Topical Clusters, Not Individual Terms

Pick a core topic your business actually owns. Then build out every subtopic, question, and comparison a reasonable buyer would ask. Tools like Ahrefs, Semrush, and Google Search Console surface these, but the real work is judgment: which of these can we actually answer better than what already ranks? Volume alone does not answer that. I look at the quality of existing results first. If the top pages are thin, outdated, or written by people with no firsthand experience, there is an opening.

  • Use People Also Ask boxes to find question-format gaps your competitors have not addressed
  • Check the existing SERP before committing — if results are dominated by large brands or aggregator sites, assess whether you have a realistic angle of differentiation
  • Prioritize low-competition, high-intent terms when you are still building domain authority
  • Long-tail clusters often convert better than head terms — a 150-search-per-month question from someone deep in a buying decision is worth more than a 5,000-search informational term

Understand Search Intent Before You Write Anything

Search intent is the most misunderstood part of this process. A keyword like schema markup examples pulls informational results. Schema markup service pulls commercial ones. Building a page for the wrong intent means Google will not rank it regardless of how well-written it is. I have seen clients spend months producing content that should have ranked, and the problem was simply that they built a service page where the SERP wanted a tutorial.

  • Informational intent: how-to guides, definitions, what-is explainers
  • Commercial investigation intent: comparisons, best-of lists, reviews, alternatives pages
  • Transactional intent: hire, buy, get a quote, pricing pages
  • Navigational intent: brand-specific and branded-product queries

Matching format to intent is not optional. Google has gotten very good at detecting when a page serves the wrong purpose for a query, and it simply does not rank it — no matter how authoritative the domain.

Competitive Gap Analysis: Where to Find the Real Opportunities

Once I have a target topic mapped out, I pull the top five ranking pages for the primary term and note every subtopic, subheading, and question they cover that my client does not. That gap list becomes the first draft of the content brief. But I also look for the reverse: topics the client covers in depth that none of the ranking pages address well. Those are differentiation angles — places where publishing something genuinely better has a realistic shot at displacing a weaker incumbent.

The tool workflow I use most: export the ranking URLs into Ahrefs Content Gap, run it against the client domain, and sort by keywords where competitors rank in the top ten and the client ranks nowhere. Filter that list down to terms with clear business relevance. What remains is usually a prioritized backlog that can feed six to twelve months of content planning without guessing. The key is not just finding gaps — it is finding gaps where you can plausibly win and where ranking would actually send the right kind of visitor.

Keyword Cannibalization: When Your Own Pages Compete Against Each Other

Cannibalization is one of the most common structural problems I find on sites that have been producing content for a few years without a clear architecture. You end up with three posts that all target “SEO for small business” in slightly different ways, and instead of one strong page ranking, you have three weak ones splitting authority and confusing crawlers about which URL to surface. Google often just picks one at random, which may not be the page you would choose.

The fix is not always to delete pages. Often the right move is to consolidate two or three thin posts into one authoritative piece, redirect the old URLs to it, and update internal links accordingly. In cases where the pages serve genuinely different intents — one is a definition post, one is a how-to, one is a service page — keeping them separate but differentiating them clearly at the content level can work. The diagnosis has to come before the prescription. I have seen consolidations that doubled traffic on the surviving page within sixty days, and I have also seen unnecessary merges hurt pages that were distinct enough to stand alone.

Prioritization Is Where Most Teams Stall

You will always have more keyword opportunities than bandwidth. I filter by three things: business relevance (does ranking here actually bring the right visitor?), ranking feasibility (do I have the authority to compete given my current site standing?), and content cost (how much effort does a truly competitive page require?). A term with 200 monthly searches and low competition often outperforms a 2,000-search term that you have no realistic shot at in the near term. Chasing the big numbers without that feasibility filter is one of the most common ways SEO budgets get wasted.

When Keyword Research Misleads You

The honest counter-case: keyword research tools report historical data, and search behavior shifts. A term showing 1,000 monthly searches in a tool may be declining, seasonal, or dominated by a feature snippet that earns zero clicks. I have seen clients invest in content targeting terms that looked great in the tool but generated almost no traffic when they ranked, because the SERP was structured in a way that answered the question before anyone clicked through. Always check the actual SERP for click-through patterns — if the top results are all rich snippets, knowledge panels, or featured answers, the organic click pool may be much smaller than the volume number suggests.

Ongoing Research Is Not Optional

Search behavior shifts. New questions emerge. Competitors move into your territory. I review keyword data quarterly for every active engagement. Search Console is the most underused tool in most clients setups — the Performance report shows you what you already rank for, which is often the fastest path to traffic gains. A page sitting at position 8 or 9 for a relevant term is usually a faster opportunity than launching an untested new topic. I look at impression data for pages that already have traction, identify what additional questions those pages could answer, and expand them before building new ones.

If you want a structured keyword research process built around your specific market, I cover this as part of my SEO consulting engagements. It is also worth reading how this work connects to writing effective meta descriptions that turn rankings into actual clicks.