Keyword research is the foundation of every SEO campaign. Get it wrong and you'll spend months creating content and building links for terms that will never drive revenue. Get it right and you have a clear roadmap for where to focus your effort.
The process I'm sharing here is the exact system we use at WEBRIS for every client engagement. We've refined it over years of running SEO campaigns for B2B companies, and it's designed to do two things: find the keywords your existing pages should target, and identify gaps where new content will drive the most impact.
In 2026, the keyword research process has evolved. AI tools can accelerate the data gathering, but the strategic thinking -- choosing the right keywords, understanding intent, and prioritizing by opportunity -- still requires human judgment. Here's the complete process.
The keyword research process at a glance
Before we dive into the details, here's the high-level workflow. Every keyword research engagement follows these five steps, in this order:
Step 1: Find the main keyword for every page
Start with Google Search Console. Export the performance data for your client's site -- you want pages, queries, clicks, impressions, and average position. This tells you what Google already associates with each page on the site.
For every page, you're looking for the "main keyword" -- the primary search term that page should target. Usually this is the query with the highest impression count that's actually relevant to the page's content. Sometimes the data surprises you. Clients often think they're targeting one keyword when Google is actually showing them for something completely different.
Cross-reference GSC data with Ahrefs or Semrush to get search volume and keyword difficulty scores. The main keyword should have meaningful search volume, clear relevance to the page, and be something the page has a realistic shot at ranking for.
In 2026, you can feed your GSC export to Claude and have it identify the best main keyword candidate for each page in seconds. But always review the output -- AI is great at pattern matching but doesn't understand your client's business priorities the way you do.
Step 2: Find secondary keywords
Secondary keywords are the long-tail and supporting terms that make a page contextually rich. These aren't synonyms of the main keyword -- they're related searches that give Google more signals about what the page covers.
Two reliable sources for secondary keywords: First, search Google for your main keyword and look at the top 3 organic results. Plug each URL into Ahrefs to see all the additional keywords those pages rank for. Those are your seedling secondary keywords. Second, look at Google's "People also ask" and related searches at the bottom of the results page. These tell you exactly what searchers want to know about your topic.
For example, if your main keyword is "personal injury lawyer Miami," your secondary keywords might include "best PI attorney in Miami," "how much does a personal injury lawyer cost," and "personal injury settlement timeline Florida." These become H2 headers and content sections within the page.
Don't confuse secondary keywords with semantic keywords. You're not looking for vaguely related terms -- you're looking for specific search queries that someone might use when researching the same topic as your main keyword.
Step 3: Map keyword intent
Every keyword carries intent -- what the searcher actually wants when they type that query. Getting intent wrong means creating content that Google won't rank, no matter how good it is.
We classify intent into four stages:
Awareness: Broad, early-stage queries where the searcher is just learning about the topic. Example: "what is SEO." These map to educational blog content.
Discovery: More specific queries where they're exploring solutions. Example: "SEO strategies for law firms." These map to in-depth guides and how-to content.
Consideration: Pre-purchase queries comparing options. Example: "best SEO agency for personal injury firms." These map to comparison pages, case studies, and service pages.
Transaction: Ready-to-buy queries with clear purchase intent. Example: "hire SEO agency Miami." These map to landing pages and contact pages.
The simplest way to verify intent is to search the keyword yourself and look at what Google ranks. If the top results are all blog posts, Google has determined the intent is informational -- don't try to rank a service page there. If the top results are product pages or service pages, the intent is transactional.
Step 4: Analyze the competition
For each main keyword, analyze the top 3 organic results to understand what you're up against. Pull the following data from Ahrefs for each competing page: Domain Rating (DR), the number of referring domains, monthly link velocity (how many new links they're gaining), and the content type ranking (blog post, service page, product page, etc.).
This competitive data tells you two things. First, whether it's realistic to rank for that keyword given your client's current authority. If the top 3 results all have DR 80+ and 500+ referring domains, and your client has DR 25, that keyword is a long-term play, not a quick win. Second, what type of content Google expects to see -- if all top results are 3,000-word guides, don't try to rank a 500-word service page.
We built formulas in our keyword research spreadsheet that calculate an "Opportunity Score" for each keyword based on the gap between the client's metrics and the competition's. Keywords with high search volume and low competitive gaps get the highest priority. This removes the guesswork from deciding where to focus.
Step 5: Prioritize and build the roadmap
With all the data collected, it's time to prioritize. Not every keyword deserves equal effort. Sort your keywords by opportunity score (highest first) and group them into three tiers:
Quick wins: Keywords where the client already ranks positions 5-15 and the competition is beatable with on-page optimization and a few links. These get tackled first for fast results.
Medium-term targets: Keywords with solid search volume where ranking will require new or significantly improved content plus a sustained link building effort. 3-6 month timeline.
Long-term plays: High-volume, high-competition keywords that require building authority over time. These go on the roadmap but don't get prioritized over quicker wins.
Present the findings to your client in a simple deck: here are the keywords we're targeting, here's the opportunity for each one, and here's the plan of attack organized by priority. Having this data makes it impossible for a client to argue with your strategy because the numbers speak for themselves.
Making keyword research faster with AI
In 2026, AI dramatically speeds up the manual parts of keyword research without replacing the strategic decisions. Here's where AI fits into each step:
For data gathering, export your GSC and Ahrefs data and use Claude to clean, categorize, and identify patterns across hundreds of keywords in minutes instead of hours. For intent mapping, AI can classify keyword intent in bulk with high accuracy by analyzing SERP features and content types. For competitive analysis, AI can parse competitor data and calculate opportunity scores automatically. For client presentations, AI can generate the charts, summaries, and recommendation narratives from your raw data.
The parts that still need a human: choosing the right main keyword when data is ambiguous, understanding the client's business priorities, making judgment calls about which opportunities to pursue, and translating keyword data into an actionable content strategy. AI handles the data work. You handle the strategy.
Keyword research isn't glamorous work, but it's the difference between an SEO campaign that drives revenue and one that spins its wheels on the wrong terms. Do the research upfront, prioritize ruthlessly, and let the data guide your strategy.