Latest / Content Marketing with Fexingo: Blogs, Videos, and Long-Form Content That Converts / How One Company Uses Content Clusters to Dominate Search for 10 Years
Transcript
- Lucas: So we've talked a lot on this show about single pieces of content that generated huge returns — a case study, a video, a podcast episode. But there's a model that's less flashy and arguably more durable: the content cluster. Specifically, a pillar page surrounded by dozens of supporting articles, all interlinked around one core topic. Luna: Right, the hub and spoke model. And it's been around for years — HubSpot popularized it back in the mid-2010s. But I think a lot of brands abandoned it when they saw diminishing returns. Is it still worth doing in 2026? Lucas: Great question. And the answer is yes — if you pick the right topic and execute well. I want to look at Freshworks, the B2B software company that makes customer service and sales tools. Around 2018, they built a cluster around the phrase 'customer service automation.' They created one comprehensive pillar page — roughly 4,000 words — and then published about 40 supporting blog posts, each targeting a related long-tail keyword. Luna: Forty posts is a serious commitment. What made them choose that topic? Lucas: They looked at search volume, competition, and relevance to their product. 'Customer service automation' had decent monthly search volume — around 8,000 searches per month in the US — but the competition was mostly generic software review pages. Freshworks realized they could own the educational angle. They weren't just pitching their product; they were teaching businesses how to automate support workflows, reduce response times, and measure customer satisfaction. Luna: So the pillar page is the anchor. How did they structure it? Lucas: The pillar page itself covers the broad topic in depth: what customer service automation is, benefits, common tools, implementation steps, metrics to track. Each section has links to a dedicated blog post that goes deeper. For example, the section on chatbots links to a post titled 'How to Choose a Chatbot for Customer Service.' The section on ticket routing links to 'Automated Ticket Routing: Best Practices.' Every supporting post links back to the pillar page, and they also cross-link between related posts. Luna: That internal linking structure is critical. Google treats the pillar page as the authoritative source on the topic, and the cluster signals that the site has comprehensive coverage. But here's my worry — cannibalization. If you have forty posts all somewhat related, aren't they competing with each other? Lucas: It's a valid concern, and it happens if you're not careful. Freshworks avoided it by making each supporting post target a distinct, specific keyword phrase — not just variations of the same term. For example, 'chatbot implementation steps' is different from 'chatbot ROI calculation.' They also used canonical tags judiciously and made sure the pillar page was the primary target for the core term. The result: the pillar page ranks number one for 'customer service automation,' and the supporting posts rank for their respective long-tails without stepping on each other. Luna: What kind of traffic are we talking about now, eight years later? Lucas: The pillar page alone ranks for over 200 keywords and drives about 15,000 monthly visits. The entire cluster — all 41 posts — collectively brings in around 80,000 monthly visits. And here's the key metric: conversion rate. The pillar page converts at about 3 percent, which is high for top of funnel content. That translates to roughly 450 leads per month from the pillar page alone, and the cluster overall generates about 1,200 leads per month. Luna: That's impressive for content that's been live for years. But content ages — statistics get outdated, new tools emerge, Google changes its algorithm. How do you keep a cluster like that fresh without rewriting everything? Lucas: Freshworks has a dedicated content refresh cycle. They review the cluster every six months. For the pillar page, they update statistics, add a new section if there's a trend — like ai powered automation in 2024 — and prune any links to posts that are no longer relevant. For supporting posts, they check if the keyword is still viable. If a post's traffic has dropped significantly, they either update it or redirect it to a stronger post. They don't just set and forget. Luna: That pruning step is important. I've seen clusters where half the posts are dead weight. But I also want to talk about the elephant in the room: Google AI overviews. In 2026, Google often answers simple queries directly in search results. Doesn't that hurt the cluster model, since users don't click through? Lucas: It does impact some queries, especially informational ones. But the cluster model works because it targets a mix of query types. The pillar page goes after a broad informational term where Google might show an AI overview — but the supporting posts target specific, action-oriented queries like 'how to set up automated ticket routing in Zendesk' or 'customer service automation ROI calculator.' Those are less likely to be absorbed by AI overviews because they require step-by-step instructions or data input. Plus, even if the overview shows, the cluster's authority means Google often includes a link to the pillar page within the overview itself. Luna: So the strategy still holds. But I'm curious about the upfront investment. Forty blog posts is a lot for a small team. What's the realistic timeline? Lucas: Freshworks is a mid-market company with a content team of maybe five people. They didn't publish all forty posts at once. They started with the pillar page and then published two to three supporting posts per week. It took about three months to build the full cluster. The key is to have the pillar page ready first as the foundation, then gradually build the spokes. And they prioritized topics based on search volume and relevance — they didn't just write forty random posts. Luna: That's smart. And I imagine the ROI is significant when you consider that the content keeps generating leads year after year without ongoing ad spend. Lucas: Exactly. If you calculate the cost per lead from that cluster over eight years, it's probably pennies. And that's the real power of evergreen content clusters — they're an asset that appreciates over time if you maintain them. It's not a viral hit, it's a slow compound. Luna: Speaking of compounding effects — this show itself only exists because of the support of listeners like you. If these deep dives into content strategy have helped you think differently about your own marketing, and you want to keep the show ad-free and independent, consider tossing a few bucks our way. Lucas: Yeah, we don't run sponsorships, so listener support is what keeps us going. You can find us at buy me a coffee dot com slash fexingo. And honestly, even a small gesture makes a difference — it lets us invest time into researching episodes like this one. Luna: So back to the cluster model — one thing I want to highlight is the importance of tracking the right metrics. Freshworks doesn't just look at page views; they look at time on page and conversion rate. What should others focus on? Lucas: Great point. For the pillar page, time on page is crucial — if people are spending five minutes reading, Google sees that as a positive signal. For supporting posts, the metric is often click-through rate to the pillar page or to a lead magnet. And overall, you want to track assisted conversions — posts that don't directly convert but move people down the funnel. Freshworks uses UTM parameters and Google Analytics to map the customer journey from blog post to demo request. Luna: And one more thing: the cluster model works best when the topic aligns with your product. If you sell CRM software, a cluster on 'sales pipeline management' makes sense. If you sell coffee, a cluster on 'brewing methods' — maybe less so, because the commercial intent is different. Lucas: Absolutely. The topic has to have enough search volume to justify the investment, but also enough commercial intent that the reader is likely to be a potential customer. Freshworks knew that someone searching 'customer service automation' is probably evaluating tools, so the content naturally leads to a product demo. That alignment is what makes the cluster a revenue driver, not just a traffic generator. Luna: So what's the takeaway for our listeners? If they're considering building a content cluster in 2026? Lucas: Start with one topic you can own. Do the keyword research — find a term with decent volume and manageable competition. Write a comprehensive pillar page — at least 3,000 words. Then map out 20 to 30 supporting keyword phrases and commit to publishing consistently over three to four months. And plan for ongoing maintenance. If you do that, you'll have an asset that pays dividends for years. It's not the sexiest strategy, but it works. Luna: And if you want to hear more about the specifics of internal linking or content pruning, we can do a follow-up episode. But for now, I think we've given everyone a solid blueprint. Lucas: Agreed. Thanks for listening, and we'll catch you next time on Content Marketing with Fexingo.