Keyword Research

Keyword research is the process of finding the searches your buyers make and deciding which ones are worth a page. It weighs each search on volume, difficulty and, above all, whether the person searching could become a client. It decides which pages you build and what each page needs to answer.

The searches worth most to a B2B company often have almost no volume and almost no competition. In Ahrefs’ US data from October 2026, “SD-WAN” gets about 6,900 searches a month, has a keyword difficulty of 47, and shows an AI Overview. “SD-WAN for healthcare” gets about 40 searches, has a difficulty of 0, and was the only one of the 12 SD-WAN and managed detection terms we checked with no AI Overview. For MSSP clients I build pages like that on purpose: one service, one industry, written for the ideal buyer. They never bring big traffic. They bring the right visitors, and those visitors request demos and network audits.

What the numbers look like across the buying journey

Keyword Stage Monthly searches (US) Difficulty Cost per click AI Overview
SD-WAN Category 6,900 47 $4.50 Yes
What is MDR Category 2,000 2 $8.00 Yes
MDR vs EDR Evaluation 900 4 $12.00 Yes
SD-WAN vs MPLS Evaluation 350 0 $5.00 Yes
Managed SD-WAN services Decision 250 7 $15.00 Yes
MDR pricing Decision 300 2 $18.00 Yes
SD-WAN for healthcare Decision 40 0 None No

Source: Ahrefs Keywords Explorer, US, October 2026. Ahrefs estimates volumes; treat them as relative, not exact.

Cost per click is the most underused number in that table. Advertisers pay for clicks that turn into revenue, so a high cost per click tells you buyers search that term. “Managed detection and response” gets about 3,500 searches a month at roughly 70 cents a click. “MDR pricing” gets 300 at $18. The smaller term is the one advertisers have learned is worth paying for.

The three layers of B2B keyword research

  • Category terms such as “SD-WAN” or “what is MDR.” High volume and mostly informational. They build topical authority but rarely produce leads directly, and nearly all of them now show AI Overviews.
  • Evaluation terms such as “MDR vs EDR” or “SD-WAN vs MPLS,” where buyers compare options and vendors. These feed comparison pages.
  • Decision terms such as pricing, managed services and industry-specific searches, where the searcher is close to choosing a provider.

Most B2B SEO programmes put their effort into the first layer because that’s where the volume is. The second and third layers produce the leads.

Why low-volume keywords look smaller than they are

Every volume number you see is an estimate, and the estimates are weakest at the low end. Google’s Keyword Planner shows ranges such as 10 to 100 instead of exact figures to advertisers with lower spend, a change Google confirmed to Search Engine Land in 2016. Search Console leaves out some queries entirely: Google’s help page says some are omitted to protect user privacy. Ahrefs says these anonymised queries hide nearly half of a site’s keyword data on average. Niche B2B searches are exactly the kind most likely to be hidden. A keyword that shows zero volume can still bring in a buyer every month.

Keyword research for AI search

Buyers don’t type keywords into AI tools the way they type them into Google. G2’s 2026 AI Search Insight Report found 33% of software buyers open their AI research with a category comparison question, and 31% start by naming competitors. Google’s documentation says AI Overviews and AI Mode may split a question into subtopics and search for each one. Your keyword list now needs the questions around each keyword, not just the keyword. And ranking alone won’t get you into AI answers: Ahrefs found only 12% of links cited by ChatGPT, Gemini and Copilot rank in Google’s top 10 for the same prompt, on average.

What keyword research produced at BroadConnect

At BroadConnect Canada, a business phone and internet provider, keyword research in Ahrefs between 2016 and 2019 focused on the commercial searches buyers used, such as “phone systems” and “business internet.” The site reached the top three for 200 of those keywords, ahead of Rogers, Telus and Bell, and its first-page keywords grew from 122 to 463, measured in Ahrefs.

Common mistakes

  • Ranking keywords by volume first and buying intent second.
  • Discarding keywords that show zero or low volume in tools that can’t measure niche searches well.
  • Ignoring cost per click, the clearest public signal of which searches buyers make.
  • Building one page per keyword instead of one page per topic and its questions.
  • Researching only Google keywords and never checking what buyers ask AI tools.

Related terms

Search intent · Search intent for B2B · Topical map · Content gap · Bottom of funnel content

Building service and industry pages for a security provider? See how I approach SEO for MSSPs.

Let's build a AI SEO search strategy that actually delivers.

Book a consult to see where your current search visibility stands, on Google and across AI search, and what is actually possible from here.