Introduction

Ranking well in search isn’t just about targeting individual keywords. It’s about building authority around topics that matter to your audience and your business.

Topic clustering is my preferred approach. It organizes related content around a central topic. This helps you better cover user intent, strengthen connections between pages, and make your content easier for people and search engines to navigate.

That foundation matters even more as search behavior changes. Google’s AI Overviews and AI Mode can synthesize information across multiple queries and sources. At the same time, more people are using tools like ChatGPT, Gemini, Grok, and Perplexity to find answers.

A strong topic cluster needs to do more than help individual pages rank. It should cover the related questions, intents, and subtopics that search engines and AI systems may use to understand and answer broader queries.

This guide will show you how to research, build, and connect topic clusters around your audience and what you sell. It will account for both search intent and AI answer behavior.

It will also cover common mistakes that can weaken a cluster. These include poor internal linking, gaps in topical depth, choosing the wrong content approach, failing to promote your content, and letting clusters become outdated.

Why topic clusters carry more weight in AI search

Google introduced a “topic authority” system in May 2023 for news, though SEOs had discussed topical authority for years. In March 2024, it folded its helpful content signals into core ranking.

AI Overviews and AI Mode use those same core ranking and quality systems. From Google’s perspective, AEO and GEO are “still SEO”.

Both features may also issue “multiple related searches across subtopics and data sources” to build an answer. Google calls this query fan-out. Those subtopics are the same ones a cluster plan starts from.

Pages that ranked for the main query and at least one fan-out query were 161% more likely to be cited in AI Overviews than pages that ranked only for the main query.

So I focus on depth: covering recurring subtopics instead of building a page for every fan-out.

On informational queries, brands cited in an AI Overview earned about 120% more organic clicks per impression than uncited brands on the same SERPs.

Getting cited helps, but it doesn’t restore old click rates. Cited brands still received 38% fewer clicks per impression than queries without an AI Overview. So I plan traffic targets around that lower number.

Research the cluster around what you sell

Let’s say a pet brand sells grooming products and wants to build a cluster under dog care.

Start by mapping the topic: your core product and every related subtopic. I still use Google Trends, Keyword Planner, Ahrefs or Semrush, and competitor site architecture. Competitive analysis is important here.

Next, research your customers and what they want to know at each stage. Keyword tools help, but so do community discussions. I look at Reddit threads, X, and YouTube comments to see the language people use when nobody is selling to them. Support and sales call notes are useful, too.

I also run buyers’ questions through AI Mode and ChatGPT. I note which sub-questions each answer covers and which sources it cites.

Why ‘best dog breeds for apartments’ is the wrong spoke

Our pet brand could try to rank for “best dog breeds for apartments.” It might pull traffic, but how many of those shoppers are ready to buy dog shampoo? In a hyper-competitive space, that may be too Top of Funnel.

It’s better to drill down and own that niche.

Design hubs and spokes by intent and answer behavior

My preferred structure is still hub-and-spoke. For our example, dog care is the hub, with spokes for grooming, exercise, gear, and nutrition.

Each spoke gets pages for every funnel stage. Informational comparison queries triggered an AI Overview 95.4% of the time.

Wire the cluster into your site

I link the hub to every spoke, and each spoke back to the hub using descriptive anchors. After adding the links, I run Screaming Frog on the cluster folder.

Measure the cluster against pipeline, then refresh it

Report at the cluster level. Group URLs by folder or regex so the grooming cluster appears as a single line item in each report.

Start with one cluster you can tie to revenue

Pick the cluster closest to revenue and give it a quarter.

Conclusion

In the evolving landscape of SEO, especially influenced by AI, it’s essential to develop topic clusters that are comprehensive and relevant to your audience's needs. This approach not only enhances search rankings but also aligns content more closely with user intent.