Lesson 5: Keyword Research That Still Works

Keyword research hasn't fundamentally changed. You're still trying to find queries where you can create the best answer. But the bar for "best answer" is higher now, which means your keyword selection matters more.

The goal isn't finding keywords with high search volume. It's finding keywords where you can realistically rank, and where ranking actually helps your business.

Start With Customer Language

The best keyword research starts offline. What questions do your customers actually ask? What language do they use when describing their problems?

Check your sales calls, support tickets, emails, and chat logs. The exact phrases customers use are often better keywords than anything a tool suggests. Customers don't search for "B2B SaaS content marketing automation platform." They search for "how to write blog posts faster" or "AI tool for marketing teams."

If you don't have customer conversations yet, look at where your audience hangs out. Reddit threads, Slack communities, forum discussions, competitor review sites. Real people describing real problems in their own words. That's your keyword seed list.

The Keyword Viability Test

For every keyword you consider, ask three questions.

First: can you be the best answer? Look at what currently ranks. If it's comprehensive guides from massive sites with huge authority, you probably can't beat them on a head-to-head comparison. But if the current results are thin, outdated, or missing a key angle, that's your opportunity.

Second: does ranking actually help your business? Some keywords drive traffic that never converts. "What is content marketing" might have great volume, but someone searching that is probably not ready to buy your content marketing tool. "Best AI writing tool for small marketing teams" is lower volume but much closer to a buying decision.

Third: can you cover it credibly? If you're writing about something outside your expertise, it'll show. Stick to keywords where you have genuine knowledge or can develop it through research and experience. Trying to rank for things you don't really understand is a waste of time.

Finding Gaps the AI Content Missed

AI content tends to cluster around obvious, high-volume keywords. That's where the tools point, so that's where everyone goes. The opportunity is in the spaces between.

Long-tail keywords (longer, more specific queries) often have less competition and clearer intent. "How to do keyword research" is competitive. "How to do keyword research for a B2B SaaS startup" is less competitive and more specific about who you're helping.

Question keywords are another gap. "What," "how," "why" queries often have poor coverage because they're harder to monetize directly. But they're how real people search, and they're often underserved.

Comparison and alternative keywords can be goldmines. "[Competitor] vs [Competitor]" or "[Product] alternatives" are high-intent queries where the searcher is actively evaluating options. These are harder for generic AI content to address well because they require specific product knowledge.

The Research Process

Here's a practical process that doesn't require expensive tools.

Start with your seed list from customer language. Ten to twenty phrases you know your audience uses.

Expand using Google's suggestions. Type each phrase into Google and look at autocomplete suggestions and "People also ask" boxes. These show related queries real people are searching.

Check each expanded keyword against your viability test. Can you be the best answer? Does it help your business? Can you cover it credibly? Be ruthless. A short list of winnable keywords beats a long list of aspirational ones.

Group related keywords into clusters. You'll often find five or ten keywords that are really variations of the same topic. These become one piece of content, not ten separate posts. Lesson 7 covers how to structure these clusters for maximum authority.

Prioritize by a combination of viability and business value. Your first priority should be keywords that are clearly winnable and directly connected to your offering. Informational keywords with less direct connection can come later, once you've built some authority.

Here's what this looks like for our coffee roaster:

Ten minutes of research, one focused cluster ready to write.

Tools That Actually Help

You don't need expensive keyword tools to do effective research.

Google Search Console shows you queries you're already appearing for. Often you'll find keywords where you're ranking position 10-30, close enough to improve with focused effort.

Google's autocomplete and "People also ask" reveal what people actually search. AnswerThePublic visualizes question-based keywords around any topic.

CopyJump's SEO features take this further. The Content Agent pulls real search volume and knows your sitemap, so it suggests keywords with actual demand that strengthen your existing clusters.

For competitor analysis or backlink data, Ahrefs and Semrush are the standards. Worth it at scale, not necessary to start.

Common Keyword Mistakes

Chasing volume over viability. A 50,000-volume keyword you'll never rank for is worth nothing. A 500-volume keyword where you can be #1 might generate meaningful traffic.

Ignoring intent. Keywords that look similar can have totally different intent. "Project management software" (informational/research) vs. "project management software pricing" (transactional/ready to buy). Know which one you're targeting. The next lesson on search intent covers this in depth.

Targeting keywords you can't differentiate on. If you're going to rank, you need a reason to rank. Better information, better examples, better structure, unique perspective. If you're just going to write the same thing as everyone else, pick a different keyword.

Going too broad too early. Your first content should be focused on your core topic. Expanding to adjacent topics comes after you've built authority in your main area.

What You Should Have After This Lesson

A list of 20-50 keywords that pass your viability test, grouped into clusters. You won't create content for all of them immediately. This is your backlog to pull from.

For each cluster, you should know: the primary keyword, related keywords that belong in the same content, the intent behind the search, and what angle you'll take that differentiates from existing content.

That's your foundation for the next lesson, where we'll dive deeper into intent matching.

Next: Lesson 6: Search Intent in the AI Era

Back to: Course Overview