Keyword Clustering for Content Strategy | Andy Chadwick

With Andy Chadwick, founder of Keyword Insights · Streamed live · 66 min

What this episode covers

Andy Chadwick explains keyword clustering, especially SERP-based clustering, which groups keywords by shared search results so teams can plan content around combined demand. He covers mining Reddit and Quora for zero-volume questions, a hub and spoke content model built from clusters, and using clusters to spot cannibalization. He argues tools should be used to optimize for what Google actually shows, not what an NLP algorithm predicts.

Key takeaways

  • ✦Group keywords by shared search results so many low-volume variants count as one topic with combined demand.
  • ✦Mine Reddit and Quora questions with engagement thresholds to find content gaps that standard SEO tools miss.
  • ✦Plan content as a hub and spoke structure by writing the subtopic clusters first, then the overarching guide that connects them.
  • ✦Use clusters to spot cannibalization, where multiple URLs rank for one cluster or one URL ranks for several clusters.
  • ✦Test SEO ideas on your own site first, and work on large sites to learn how complex SEO works.

Key moments

  1. 29:19SERP vs NLP clustering explained
    “we operate in Google's world as seos and content marketers we want to optimize what they're doing not what like an algorithm thinks”
  2. 33:15Reddit questions for content gaps
    “we turned to Reddit and we went through Reddit we found something really cool”
  3. 37:25Hub and spoke model explained
    “it's less about the linking and how you're linking it's more about if you plan it that way you're going to just make sure you've covered the whole topic really well”
  4. 38:57Writing clusters before the hub
    “I always write all my clusters and then I would write the overarching piece that connects them all afterwards”
  5. 52:09Cannibalization case study for property site
    “they had 420 keyword pages and what I mean by keyword Pages they had their website was 50 million URLs”
  6. 58:31Future of keyword clusters
    “we're going to stop thinking about search engine optimization and we're going to start thinking about search experience optimization”

Quotes are from the episode transcript, lightly cleaned of auto-caption errors.

Questions answered in this episode

How does SERP clustering differ from NLP clustering?
NLP clustering groups keywords by an algorithm's view of meaning, while SERP clustering groups keywords by how many URLs their search results share. Andy prefers SERP clustering because SEOs need to optimize for what Google shows.
How can you find content ideas that keyword tools miss?
Andy scraped Reddit questions filtered to posts with at least 100 upvotes and 10 comments, then clustered them so repeated questions revealed real demand. He also suggests Quora and Search Console queries as sources.
What is the hub and spoke model according to Andy?
It is one core guide that links out to focused subtopic pages, and Andy recommends writing the clusters first and the overarching hub last.

More episodes on Content & On-Page SEO