3 min read

Keyword Research Is Procrastination

You've spent three weeks in Ahrefs building the perfect keyword strategy. Meanwhile, your competitors published 20 articles and are already ranking.

I know an SEO who spent six months building a keyword research spreadsheet - 50,000 keywords, perfectly categorized, search volume and difficulty scores and business value and content gaps, the whole works, a monument to taxonomic ambition - and during that entire six-month period of feverish categorization, during which he told himself daily that he was building something essential and foundational, he published exactly zero pages.

Keyword research is the most productive-feeling way to do nothing.

The Illusion of Progress

Water Lilies by Claude Monet
Floating above the murky depths of SERP features.

Keyword research feels like work because you're in a tool, you're exporting data, you're building spreadsheets, you're "strategizing" in that peculiar way that involves lots of clicking and downloading and arranging columns - but nothing is actually happening, no pages are being created, no traffic is being earned, no rankings are being won, and you're essentially organizing your sock drawer while the house burns down around you.

The Diminishing Returns

Here's the thing about keyword research: the first 20% takes 5 minutes, and the last 80% takes months, which is exactly the sort of lopsided distribution Pareto spent his career documenting, and that first five minutes - asking yourself what questions your customers have, what problems keep them up at night, what they're typing into search boxes when they're desperate for answers - is actually enough to get started, is genuinely sufficient to begin doing the work that matters.

Everything after that initial five minutes is optimization theater - finding the "perfect" keywords, calculating opportunity scores with elaborate formulas, building content calendars that stretch into the next fiscal year - while meanwhile the person who spent those same five minutes on research and then started publishing is already ranking for long-tail variations you haven't even discovered yet because you've been too busy perfecting your spreadsheet.

The Reality of SEO

Las Meninas by Velázquez
Layers of meaning. Layers of redirects.

Here's what actually happens when you publish content: you rank for things you didn't target, because most traffic comes from keywords you never researched - long-tail variations, related queries, semantic matches that no spreadsheet would have predicted - and your research becomes obsolete almost immediately as search trends shift, new queries emerge, competitors publish their own content, meaning that perfect spreadsheet was outdated before you even finished formatting the cells.

And here's the real kicker: Google tells you what to write, because Search Console shows you what queries you're already appearing for, and you can double down on those, and actual data from real searches beats theoretical research every single time - in fact, there's a simple Search Console trick that finds quick win opportunities most people miss entirely.

The Better Approach

Spend 30 minutes on keyword research - list the obvious topics, check what competitors rank for, note any obvious gaps, and then stop, done, finished, close the tool - and then publish something, write the best piece you can about an obvious topic, make it genuinely useful to someone who actually needs it.

Check Search Console in a month to see what queries you're actually getting impressions for, what questions real people are actually asking in the wild, and then iterate based on reality - not theoretical keyword data, not projected search volumes, but actual performance from actual searches by actual humans.

This approach beats months of spreadsheet building every single time, and if you're still tempted to build that perfect keyword strategy, remember: there's no such thing as an SEO strategy anyway.

The best keyword strategy is publishing and seeing what works. The worst is perfecting a spreadsheet you'll never use.

Close Ahrefs. Open your CMS. Start writing.

Disagree? Good.

These takes are meant to start conversations, not end them.

Tell me I'm wrong