Keyword research

What is keyword research?

Keyword research is the work of finding what your buyers type into search engines and AI tools, then deciding which of those searches deserve a page on your site. It answers two questions: what to write, and in what order.

Let's say you sell contract management software. A legal ops manager might search "what is contract lifecycle management" while she's learning, "best CLM software for mid-size companies" while she's shopping and "Ironclad alternatives" once she's comparing vendors. Keyword research finds those searches, checks how many people make them and how hard they are to rank for, and turns them into a content plan.

Keyword research is a core part of my process for every B2B client. But it's easy to fall down the rabbit hole and end up with 10,000 keywords nobody will ever write for. Good research ends with a short list you can act on.

Why keyword research matters for B2B SaaS

Without it, your content plan is guesswork. You risk writing for searches nobody makes, or going after terms so hard you'll never reach page one.

B2B SaaS makes this harder, because the searches that matter most are small. A buyer-ready search like "CLM software for procurement teams" might show 50 searches a month, while a broad term shows 20,000 searches from students and job seekers.

Here's what good keyword research gives you:

• A plan sorted by revenue: You can put the searches closest to a purchase first, even when their volume looks tiny.
• Ideas you'd never think of: Tools surface terms and questions you didn't know existed, which often turn into your best pages.
• A view of the market: Seeing what people search in your niche shows you their pain points and the topics no competitor has covered well.
• Words buyers really use: Buyers rarely use your internal product names. Research tells you the words they use, so your pages match them.

How to do keyword research

I do my keyword research with SEMRush, Google and Reddit, and I start with people before tools. Here's my process:

• Map your audience first: List each buyer role, their main pain points and what your product does about it. A director of engineering and a CFO search for very different things, even for the same tool.
• Mine sales calls and support tickets: Write down the questions buyers ask in demos and the competitors they mention. These give you the BOFU searches that tools tend to undercount. Our post on how your sales team can inform content strategy shows how to set this up.
• Reverse engineer Google: Type a seed keyword into Google and see what ranks. Then plug the ranking pages into SEMRush or Ahrefs to see what other searches they rank for.
• Do community research on Reddit: Search "site:reddit.com" plus your topic in Google. The threads show the comparisons, edge cases and plain-language questions that keyword tools miss.
• Check volume, difficulty and intent: For each candidate, note the search volume, the keyword difficulty and what the results page tells you about search intent.
• Shortlist and group: Keep the searches that match what you sell, group them into topics and assign one page to each group. Then sort the list with buying searches at the top.

A keyword research example

Here's a real one from the test automation niche I used to work in. Say you want to rank for "no-code test automation software". When I Googled it, Reddit threads filled the top spots, and the rest of the results were listicles of codeless testing tools.

So I took one of those lists, from BugBug, and put it into Ahrefs to see what else it ranked for. As of October 2025, it ranked well for "codeless automation testing tools", which had a monthly volume of 330. I checked it again in SEMRush, it looked good on both volume and difficulty and it went straight onto the list. Then I repeated the same loop for the next pain point.

The format mattered as much as the keyword. Since Google was showing lists for that search, the page to build was a fair "best tools" list. I walk through the whole process step by step in my guide to B2B keyword research.

How to measure keyword research

Keyword research pays off only if the pages you build from it rank and bring buyers. Here's what I track:

• Share of target keywords on page one: Track your shortlist in a rank tracker like SEMRush and count how many reach the top 10 over six months.
• Impressions for new queries: In Google Search Console, watch which new queries your pages start to show up for. They tell you where your research was right and what you missed.
• Leads by keyword group: In GA4 and your CRM, tie demos and trials to the landing pages for each topic. A group with low traffic and steady demos is a group to expand.
• AI answers for the same questions: Run your top buyer questions as prompts in ChatGPT, Perplexity and Gemini each month. Prompt tracking shows whether the pages you built get named or cited there too.

Common keyword research mistakes

• Starting with the tool: If you begin with a tool export, you'll get thousands of terms with no link to your buyers. Start with sales calls and your audience map.
• Skipping small keywords: A search with 30 searches a month from buyers ready to switch tools can beat one with 3,000 from students. Judge keywords by intent first.
• Trusting the metrics too much: Volume and difficulty are estimates. Always look at the real results page before you commit to a keyword.
• Building one page per keyword: "CLM software" and "contract lifecycle management software" usually want the same page. Writing two leads to pages that compete with each other.
• Researching once: Buyers' words change, new competitors show up and AI tools change how people ask. Revisit the list every quarter.

Here's a first step you can take today. Open your last five sales call notes and write down every question a buyer asked. How many of them does your site answer right now?