A Topic Cluster Strategy, Six Months of Rankings Data Later

A client came to me wanting a 'topic cluster strategy' because they'd read it was how you beat sites with ten times their domain authority. I built it. It worked, mostly. But the win showed up in a different place than the sales pitch for topic clusters usually promises.

We mapped one pillar page with twelve supporting cluster pages, all interlinked with consistent anchor text, all published within six weeks of each other. The pillar page itself barely moved — it went from position 14 to position 11 over six months, which is a rounding error in traffic terms.

Where the actual lift came from

The cluster pages — the narrow, specific, long-tail pages — were the ones that moved. Four of the twelve went from unranked to top-10 within three months, and those four accounted for about 70% of the new organic traffic the whole cluster generated. The pillar page's job turned out to be organizing internal link equity, not winning its own keyword.

The pillar page is the hub for link equity, not the hero of the traffic story.

What the strategy decks don't mention

• Cluster pages compete with each other for the same internal links if you're not careful about anchor text variation
• A pillar page with twelve weak children doesn't beat a single strong page — quality per cluster page still matters more than cluster count
• The six-week publish cadence created a crawl and indexing bottleneck we hadn't planned for
• Three of the twelve cluster pages never ranked at all and should have been merged into stronger pages instead of published separately

If I ran this again, I'd publish six strong cluster pages instead of twelve adequate ones, and I'd stop treating the pillar page's own ranking as the success metric. The real metric is whether the cluster as a whole captures more total search volume than the sum of what each page could have captured in isolation — and on that measure, this one paid for itself inside the first quarter.