Why 72.9% of Top-Ranking Pages Are 3+ Years Old — And How Topic Clusters Get You There Faster

Most blog archives rot quietly.

You publish a post, it ranks for a few weeks, then it slides. Not because the post is bad, but because nothing around it tells Google it's part of something bigger. It sits there, an orphan with a title. Multiply that by two years of weekly publishing and you have a 100-post archive where each post is competing for attention instead of supporting each other. (If you're wondering whether the blog is even worth the effort anymore, the short version is yes — but the rules changed.)

Topic clusters fix that. But the way most articles explain them… it's a lot of definitions and not much doing. So consider this the operator's version: where the idea came from, what it actually looks like in 2026, the numbers behind it, and the exact moves an agency or solo operator can make this month to make their archive start compounding instead of decaying.

A quick note before we get into it. This is one of those SEO topics where everyone is selling you the same "ultimate" framework and none of them ship the hard part — the interlinking, the cadence, the updates. We'll cover that too, because the hard part is where the system earns its keep.

What a topic cluster actually is (in one sentence)

A topic cluster is a pillar page that targets a broad topic, surrounded by cluster posts that each cover a specific subtopic in depth, all linked together so search engines (and readers) can see they belong to the same idea.

The model was popularized by HubSpot back in 2017 as a reaction to the way Google was moving away from keyword-stuffed posts and toward understanding entities and relationships. A decade later, the model is still the right shape — but the execution has changed.

If you want a quick visual, the canonical HubSpot diagram is the one you'll see on every SEO blog. It works because the idea is simple: one hub, many spokes, links in both directions.

Why clusters actually work (with the data)

The honest version: Google has never officially blessed "topic clusters." The closest you'll find in Google's own documentation is the line, "Design your site to have a clear conceptual page hierarchy." That's it. The rest is SEOs (and HubSpot) building a useful framework on top of a real signal.

The signal is this: Google rewards sites that look like an authority on a topic. Authority is implied by depth, internal link structure, and the way related pages support each other. Clusters are how you show that.

Now, the data. From Ahrefs' 2025 study of 1M URLs:

• Only 1.74% of newly published pages make it to Google's top 10 within a year.
• The average #1 ranking page is 5 years old.
• 72.9% of top-10 pages are more than 3 years old.

That's discouraging if you're publishing one-off posts. It makes a lot more sense if you're publishing into a cluster that has years of supporting content and internal links. Clusters age well. Lone posts don't.

A second pattern from the same study: 40.82% of pages that did reach the top 10 did so within 1 month of publishing. Translation: when a new post is dropped into a strong cluster with good internal links, it inherits authority from the pillar and the other cluster posts. The cluster carries it.

Pillar page vs. cluster post (the real difference)

A lot of articles treat this like a vibe. It's not. They're structurally different pages that play different roles in your archive.

The pillar is your investment. The cluster posts are how you earn compounding returns on that investment over time.

A worked example helps. Say your blog is for marketing agencies. Your pillar is "The agency content engine." Around it you might build:

• "How to onboard a new client blog in 48 hours"
• "The 6-week content sprint that pays for itself"
• "Editorial calendar templates for 3, 5, and 10-client agencies"
• "How to price retainers when scope creeps"
• "The handoff doc that ends scope creep"
• "Reporting templates clients actually read"
• "How to fire a client without burning a bridge"

Eight posts. One pillar. Each one is a search someone is doing. Each one links back to the pillar. The pillar links to all of them. Six months in, the pillar is ranking for the broad term because the cluster underneath it has built up topical authority. (For a deeper look at the operational side of running multiple client blogs, see Scaling Content Operations for Marketing Agencies.)

You can see this same pattern in Ahrefs' analysis of three real topic clusters — Podia's 8-page online course guide, Wine Folly's 40+ page wine cluster, and Muscle and Strength's 700+ page workout database. Different sizes, same shape.

How many cluster posts per pillar?

This is the question every operator asks, and the honest answer is: it depends on how deep the topic goes and how competitive it is.

A few useful rules of thumb from operators who ship this stuff:

• Minimum viable cluster: 1 pillar + 5–8 cluster posts. Below this and you don't have enough internal links for the cluster to register as a system.
• Healthy mid-size cluster: 1 pillar + 15–25 cluster posts. This is where you start to see the "snowball" effect in Search Console.
• Authority cluster: 1 pillar + 50+ cluster posts. Wine Folly and Muscle and Strength territory. Only worth it if you have a system to maintain it.

You can ship the minimum viable cluster in 6–8 weeks at one post per week. You can build an authority cluster in 6–12 months. Anything slower and the cluster loses momentum because new cluster posts aren't linking to a pillar that's getting fresh signals.

Step-by-step: building a cluster that actually compounds

This is the part most articles skip. The framework. Here's the order that works in practice.

Step 1. Pick 3–5 pillars, not 30

Resist the urge to map your entire business into pillars. Pick the 3–5 problems your best customers actually have. Each pillar should be broad enough to support 10+ cluster posts but specific enough to be searchable.

For an agency, that might be: "agency content operations," "client onboarding," "agency pricing," "content strategy for SaaS." For a SaaS founder, it might be: "product-led SEO," "activation," "churn reduction," "self-serve onboarding."

If a pillar can't generate at least 10 obvious subtopics, it's too narrow. If it generates 100, it's too broad.

Step 2. Audit the orphans

Before you write a single new post, look at what you already have. Most blogs already have 10–20 posts that fit cleanly into a cluster. They just don't link to each other.

A 30-minute audit:

• Export your post URLs.
• For each post, write down the primary topic in 3 words.
• Sort by topic.
• Posts that don't have a topic — orphan posts. Either kill, redirect, or rebuild them into a cluster.

This step alone is usually worth 20–30% traffic lift on the affected posts.

Step 3. Map the cluster on paper first

Before drafting, sketch the cluster on a single page. Pillar in the center, 8–15 cluster posts around it, with the keyword each one targets.

A simple table works:

When you can see the whole cluster on one page, gaps jump out. Some clusters will have a strong hub but no posts on a subtopic that should obviously exist. That's your roadmap.

Step 4. Publish the pillar first (or rebuild it if it exists)

The pillar is the only post that has to exist before cluster posts start linking to it. If you already have a long-form post on the broad topic, that's your pillar — even if it needs a rewrite. If you don't, ship a 2,500–3,500 word pillar that covers the topic at a high level and includes placeholder links to cluster posts that will exist soon.

Don't wait for the cluster to be complete. Search engines reward the structure even if some cluster posts are still drafts.

Step 5. Ship cluster posts on a cadence

One cluster post per week is the sweet spot for most operators. Fast enough to build momentum, slow enough to ship quality. Each cluster post:

• Targets one specific long-tail keyword.
• Links to the pillar in the first 200 words and at least once more contextually.
• Links to 0–2 sibling cluster posts where the connection is genuinely useful.
• Includes a short FAQ block of 2–3 questions if the topic warrants it.

Step 6. Wire internal links, then update older posts

This is the step most operators skip. Once the cluster is in place, go back to your older posts that touch the same topic and add a contextual link to the new pillar. That single link from an old, established post is often what kicks the new pillar into ranking.

The easiest way to do this at scale: search your archive for the pillar's primary keyword, then add a link from the 5–10 most relevant older posts. This is the kind of interlinking the LotsBlog agent is built to do on autopilot — it reads your archive, finds the contextual fit, and proposes the link for your approval.

Common mistakes that kill clusters

These are the patterns that turn "topic cluster" into a content graveyard.

• Publishing a pillar with no cluster posts underneath. The pillar is just a long post. Without supporting posts linking in, it has no authority signal.
• Cluster posts that don't link back to the pillar. A cluster page that doesn't link to its pillar is an orphan with a parent it won't acknowledge.
• Using "ultimate guide" or "complete guide" as anchor text. It tells Google nothing about the linked page. Use the actual topic name.
• One pillar per site. Most blogs should have 3–5 pillars, not one. Search intent is too diverse to compress into a single hub.
• Letting the cluster go stale. If the pillar is 18 months old and you've added 15 cluster posts, the pillar needs a refresh — new sections, updated stats, links to the new clusters.
• Treating cluster count as the goal. A tight cluster of 8 strong posts beats 30 shallow ones every time. Quality and topical fit matter more than volume.
• Forgetting to measure. If you can't tell whether the cluster is working after 90 days, you'll keep doing it because it sounds right, not because it's paying off.
• Running it on a CMS that doesn't help. If your platform makes you manually link every post by hand, the cluster will decay the moment you get busy. The whole point is to make the interlinking automatic, not aspirational. (This piece on moving past WordPress for agency work goes deeper on what a system operator replaces.)

How long until you see results?

Honest answer: it depends on your domain authority, the competition, and how aggressively you build the cluster.

The Ahrefs study found that 40.82% of pages that did crack the top 10 did so within 1 month. Those are usually cluster posts that got picked up quickly because the pillar and siblings were already strong.

A more typical timeline for a brand-new cluster on a domain with some existing authority:

• Week 1–4: Posts are indexed. Some long-tail cluster posts start ranking on page 2–3.
• Month 2–3: The pillar starts climbing for its broad keyword. Internal link equity is consolidating.
• Month 4–6: Most cluster posts are ranking in the top 20. The pillar is consistently in the top 10 for the broad term.
• Month 6–12: Compounding kicks in. The cluster is now an asset: new posts you add to it inherit authority from the existing structure.

If you go six months and the pillar isn't moving at all, the topic is probably too competitive for your current authority, or the cluster isn't tight enough. That's the moment to either narrow the pillar or shift to a less competitive topic.

How to measure whether the cluster is working

Don't get fancy. Three numbers tell you almost everything:

• Cluster organic traffic: the sum of organic sessions to all posts in the cluster. Track this weekly. It should grow every month after month 2.
• Pillar keyword position: track the pillar's ranking position for its primary broad keyword. If it's stuck below position 15 after 4 months, the pillar needs more depth or more incoming links.
• Internal link clicks: in Google Analytics 4, the engagement path from cluster posts back to the pillar. If cluster readers aren't clicking through, the anchor text or the call to action isn't pulling its weight.

A good cluster hits these benchmarks by month 6:

• Cluster organic traffic up 2–3x from baseline.
• Pillar ranking in the top 10 for its primary keyword.
• 30%+ of cluster post sessions include a click through to another post in the same cluster.

A cluster is a system, and a system is something you can run from a dashboard. If you don't have one, you're flying blind. And remember: traffic is the means, not the end. If the cluster is bringing in readers who never convert, the topic may be wrong, not the structure. (The #1 reason most bloggers never make money covers that side of it.)

How to actually run this without burning out

Here's the part nobody puts in the SEO playbook. The reason most clusters die isn't bad strategy. It's that the operator runs out of energy at week 6, week 12, week 20. The pillar gets written, three cluster posts ship, and then life happens.

The pattern that breaks this is treating the cluster as a system, not a project.

A system has four parts:

1. A backlog. A running list of cluster posts in priority order, never empty.
2. A cadence. A predictable publish rhythm — once a week is the most common.
3. An automation layer. Drafting, interlinking, metadata, schema, all handled by something other than your willpower.
4. A refresh loop. A schedule for revisiting the pillar and the oldest cluster posts every 6–12 months.

If you have those four, the cluster compounds. If you're missing any one of them, the cluster will stall. This is the part that's hard to do with a CMS and an AI writing tool. It's the part that's easy to do with a system operator.

If you're running multiple client blogs, multiply the problem by the number of clients. A 5-client agency needs 5 active clusters in flight at any given time, or the agency is just trading one-off writing jobs for slightly more organized one-off writing jobs. The agency-level system is the same shape: backlog, cadence, automation, refresh — applied to a workspace of blogs, not a single one.

A 30-day plan to launch your first real cluster

If you want to stop reading and start shipping, this is the sequence.

Week 1: Decide and audit.

• Pick 3 pillars. Write them on a sticky note.
• Audit your archive. Mark each existing post with a cluster (or mark it orphan).
• Identify the 5–8 highest-value existing posts to link into the new pillar.

Week 2: Build the pillar.

• Write or rewrite the pillar page. 2,500–3,500 words. Comprehensive, but not exhaustive.
• Include placeholders for cluster posts you'll ship in the next 8 weeks.
• Add the existing 5–8 posts as contextual links from the pillar.

Week 3: Ship cluster post #1 and #2.

• Each one targets a specific long-tail keyword.
• Each one links back to the pillar in the opening 200 words.
• Add a "Related posts from this cluster" section to each.

Week 4: Ship cluster post #3 and set the cadence.

• Lock the cadence: one cluster post per week from here.
• Set up the refresh loop: a calendar reminder to revisit the pillar in 90 days.
• Add the next 8 cluster posts to your backlog so the queue never empties.

By the end of month one, you'll have a real cluster: one pillar, three to four cluster posts, an interlinking structure, and a backlog. That's more than most blogs have after a year of "we should really do topic clusters."

FAQ

Do topic clusters still work in 2026?

Yes, but the framing has changed. The model is the same — pillar + cluster posts + interlinking — but search engines now look for entity relationships and topical depth more than keyword density. A cluster that genuinely covers a topic end-to-end will outperform a flat archive of unrelated posts. The numbers haven't changed: in Ahrefs' 2025 dataset, older, deeper pages dominate, which is exactly what a well-built cluster produces.

How long does a topic cluster take to rank?

For a brand-new cluster on a domain with some authority: 3–6 months for the pillar to reach the top 10 for its primary broad keyword, and 2–4 months for new cluster posts to start ranking for their long-tail targets. Clusters on newer domains can take 9–12 months. The fastest 40% of new pages to crack the top 10 in Ahrefs' data did so within 1 month — usually because they were dropped into a strong existing cluster.

How many cluster posts do I need per pillar?

Minimum viable is 5–8. A healthy mid-size cluster is 15–25. Authority clusters go 50+. The right number depends on how deep the topic goes and how competitive the primary keyword is. Don't pad with weak posts just to hit a number — a tight 8-post cluster beats a 30-post cluster full of filler.

Topic cluster vs. content silo — what's the difference?

A content silo is an older model from the 2000s that focused on URL structure and directory-based hierarchies. A topic cluster is a more modern model focused on internal link relationships and topical depth. In practice the two converge: both group related content together, both emphasize internal links, both aim to build authority. Topic clusters are the dominant model now because they map better to how Google actually understands relationships between pages.

Can a single site have multiple pillars?

Yes — most sites should have 3–5 pillars, not one. Search intent is too diverse to compress into a single hub. Each pillar represents a major problem your audience has, and each one becomes the center of its own cluster. The pillars should be loosely related (they all sit under your broader topic) but not overlap.

Should I 301 old posts into the pillar?

Generally no. Most old posts have their own long-tail value and bring in their own traffic. The better move is to keep them live, add a contextual link to the new pillar, and let the internal link signal do the work. A 301 is only worth it if the old post is truly duplicate content with no unique value — and even then, a 301 to a cluster post often performs better than a 301 to the pillar, because the cluster post is a closer topical match.

How does LotsBlog fit into this?

LotsBlog runs the cluster as a system, not a project. Each blog comes with an agent that plans the next post based on what's in the archive, drafts it, links it to the right pillar and siblings, schedules it, and revisits older posts when facts change. It's the same agent-driven architecture that powers LotsAgent, applied to the content lifecycle. So instead of running the cluster out of a spreadsheet, you're running it out of a workspace.

Run the cluster as a system

A topic cluster is an operating model, not a content strategy. The model is the same whether you're a solo operator or an agency running 30 client blogs — what changes is whether the system is set up to keep it compounding.

Spin up a free blog on LotsBlog and point the agent at your archive. It plans the next post, drafts it, links it into the right cluster, and proposes the publish — you approve what ships.

Start your free blog →

If you're running the same cluster pattern across multiple client blogs, the workspace scales the same way: one operator, many blogs, one agent per blog.

The cluster is the strategy. The system is what makes the strategy ship.