Sorting Ahrefs by volume and picking the top results is not keyword research. Modern research is intent modelling, SERP feature analysis, opportunity scoring and topical clustering. Here's the workflow our strategists use when scoping new accounts.
1. Universe Construction
Start with seed terms from your existing rankings, competitor rankings (Ahrefs gap), AI Overview surface area and audience research. Don't filter yet — capture everything. Typical universe size for a B2B SaaS: 8,000–15,000 candidate keywords.
2. Intent Classification
Classify every term by intent — informational, commercial investigation, transactional, navigational. The SERP itself is the truth here: what types of results does Google rank? Tutorials? Comparisons? Product pages? Classify based on what's currently winning, not what you'd prefer.
3. SERP Feature Analysis
Featured snippets, AI Overviews, People Also Ask, image packs, video carousels — every feature compresses the click-through opportunity. Score each term by SERP feature density. A high-volume term with three SERP features sitting above the organic results is worth less than a lower-volume term with clean organic real estate.
4. Difficulty vs Opportunity
Don't trust the difficulty score in isolation. Combine it with your domain's current authority profile, content gap analysis and competitor weakness scoring. A "DR70" term might be reachable for a DR50 site if the ranking competition is weaker than the domain rating suggests.
5. Business Value
The most important filter, applied last. Tag every keyword with funnel stage, expected conversion behaviour and an estimated revenue contribution. This is the layer most agencies skip — and it's why so many SEO programmes deliver "traffic" that never becomes pipeline.
6. Topical Cluster Mapping
Group the final list into pillar/cluster relationships. Identify which pillars your domain can credibly own. Build the content map from the cluster, not the keyword.