Topic clusters: fixing the cannibalization trap

September 6, 2026 · 5 min · Free, with full transcript

Why does publishing more content often produce less traffic? The answer here is keyword cannibalization: twenty half-finished posts on the same subject split the credit between them and none of them ranks. The position taken is that authority is topical, not page by page, and that a pillar page supported by interlinked cluster pages is the structure search engines read as depth. You get the fintech audit that merged nine pages on how to build credit into one guide and tripled topic traffic in four months, the HubSpot interlinking experiment, Ahrefs as a worked example, why AI-generated search answers raise the value of depth, and the two failure modes: missing internal links and never pruning near duplicates.

Listen to the podcast

4 min

Chapters

Key takeaways

  • Sort pages by the search query bringing them traffic and find two or more ranking for the same phrase.
  • Merge weak duplicate pages into the strongest one and redirect the old URLs to consolidate authority you already earned.
  • Build one broad pillar page per subject, then narrower cluster pages that link back to the pillar and to each other.
  • Pick three topics you can cover obsessively rather than spreading thin across a hundred.
  • Treat a cluster like a garden, pruning and merging pages, instead of a factory that keeps adding near duplicates.
Read the full transcript

Welcome to the Leaders Insights podcast.

Today's episode, topic clusters and pillar content, the architecture behind organic authority.

Most content teams publish more and rank less.

Why does volume so often backfire?

Because they build a pile, not a structure.

20 blog posts, all half covering email marketing.

And Google has no idea which one to trust.

So it splits the credit across all of them, and none of them wins.

That's called keyword cannibalization.

Your own pages eating each other's rankings.

That sounds like a self-inflicted wound.

How common is it, really?

Terrifyingly common.

When we audited a fintech client in early 2026, they had 11 separate pages targeting variations

of how to build credit.

11.

Every one of them sat on page 3.

We merged 9 into 1 authoritative guide, and traffic to that topic tripled in 4 months.

Same content, roughly.

Just stopped competing with ourselves.

So the fix is topic clusters.

Give me the plain English version, because it sounds like consultant jargon.

It's an architecture.

You pick one broad subject and write a big central page.

The pillar that covers the whole territory.

Then you write narrower pieces.

The cluster pages.

Each digging into one sub-question.

And you link them all back to the pillar and to each other.

Search engines read those links as a map.

They see you've covered the subject with depth, not just dropped one lucky article.

Isn't that just internal linking with a fancy name?

Internal linking is the plumbing.

The cluster is the building.

The difference is intent.

Random links say nothing.

A deliberate structure says, we own this topic.

HubSpot.

And yes, they sell content software.

So they've got a horse in this race.

They ran their own experiment years back and found that the more interlinked a cluster was,

the higher every page in it ranked.

Not the pillar alone.

The whole neighborhood rose together.

Fine, but that's a durable claim.

What actually changed in 2026 that makes this matter more, not less?

The rise of AI-generated answers in search.

When a machine summarizes the web to answer a question directly, it isn't grabbing your

one clever post.

It's assessing whether you're a credible source on the entire subject.

Depth is now the moat.

A single article is a lottery ticket.

A cluster is a reputation.

And reputations are what gets cited.

So the operators who figured this out early, what did they actually internalize that everyone

else missed?

That authority is topical, not page by page.

Amateurs chase individual keywords.

The good ones ask, what subject do we want to be the obvious expert on?

And then build every piece to reinforce that one answer.

RFS, the SEO tool company, is a master class.

Nearly everything they publish points back to a small set of pillar guides.

You can't research their subject without tripping over them.

That's not luck.

That's design.

Devil's Advocate.

Doesn't this reward big brands with huge content budgets and crush the small player?

Backwards, actually.

Clusters favor the focused.

A small brand can't outpublish a giant across 100 topics, but it can absolutely dominate

three.

Pick a narrow territory, cover it obsessively, and you'll outrank a bloated competitor who

spread themselves across everything.

Constraint is the weapon here.

The big brands publish wide and shallow.

You publish narrow and deep, and depth wins the subtopic every time.

Where do people botch it?

Because just build clusters sounds too clean.

Two ways.

First, they build the cluster and forget the links, so it's a structure with no wiring,

invisible to the machine.

Second, they never prune.

They keep publishing new near duplicates and recreate the cannibalization they were trying

to fix.

A cluster needs a gardener, not a factory.

Sometimes the highest value action is deleting or merging three weak pages into one strong

one.

Give me the one thing a listener can do Monday morning.

Open your analytics.

Start your pages by the search query, bringing them traffic.

And look for two or more pages, ranking for the same phrase.

Pick the strongest.

Merge the rest into it.

Redirect the old links.

You'll consolidate authority you already earned and probably watch that survivor climb within

weeks.

Stop competing with yourself before you build anything new.

Fix the plumbing before you extend the house.

That'll do it.

That's a wrap.

Fresh CMO briefings drop daily at MBA-training.com.

Thanks for watching.

See you next time.

Bye.

Read and go deeper