HubSpot: Old Posts That Doubled Leads

Case study · Marketing 465 · Analytics with GA4

HubSpot’s blog team looked at its analytics and found that most of its traffic and nearly all of its leads came from posts it had already published. So it started fixing old posts instead of only writing new ones, and leads from those posts more than doubled.

The situation

By 2014, HubSpot ran one of the biggest marketing blogs on the internet. The team published about 200 new posts a month and had close to 6,000 posts in its archive. Like most content teams, it measured success by what it shipped next.

Then the team pulled its own blog analytics and looked at where views and leads actually came from. The answer surprised them. Most of the action was not on the new posts at all.

The challenge

The numbers showed that about three of every four monthly blog views, and more than nine of every ten monthly blog leads, came from posts published in earlier months. Just 30 old posts drove almost half of the leads. The team had been pouring effort into new content while its best performers sat untouched.

Many of those older posts were also going stale. Facts aged, screenshots went out of date, and the calls to action did not match what readers were searching for. The question became: what if the team treated its archive as an asset to manage, not a pile of finished work?

What they did

HubSpot started a project it called historical optimization. Instead of writing only new posts, a small part of the team’s time went to finding old posts with real search demand and making them better. David Meerman Scott wrote about the project in late 2014, when it was just getting started.

The work followed a simple routine that any team can copy.

• Find the winners: export blog analytics and sort old posts by views and leads.
• Match the offer to the search: check which keywords each post ranks for, then rewrite its call to action to fit that search.
• Update the content: fix outdated facts, add depth and make the post the most complete answer to the query.
• Republish and promote: give the post a new date, then push it through email and social like a new post.
• Keep it steady: the team updated about two or three posts a week, not everything at once.

The key move was looking at the data first. The plan came from what the analytics showed, not from a hunch about what readers wanted next.

What happened

• Monthly leads from the optimized old posts more than doubled (HubSpot).
• Monthly organic search views of the updated posts rose 106% on average (HubSpot).
• Keyword-matched calls to action were added to 75 posts, and all 75 saw higher conversion rates (HubSpot).
• Late 2024: HubSpot’s organic traffic reportedly fell from about 13.5 million to 8.6 million in a month, mostly on the blog, after Google updates and AI Overviews (Search Engine Land, using Semrush data).
• 2025: HubSpot shifted part of its search work toward being cited in AI answers, and a former leader reported AI-sourced qualified leads up 1,850% (Growth Unhinged, 2026).
• The catch: the results are HubSpot’s own numbers, and the company sells marketing software that supports this kind of work.

Why it worked

The analytics changed the question. The team stopped asking “what should we write next?” and started asking “where do our leads really come from?” One honest look at the data showed the archive was doing most of the work.

Small fixes, big base. A post that already ranks and already gets traffic only needs to get a little better to pay off. Improving the call to action on a popular post can beat writing ten new posts that nobody finds.

It was a routine, not a stunt. Two or three updates a week is a pace a real team can hold. That steady rhythm let the gains stack up month after month.

Where it gets complicated

• Changing the date is not the trick. Google’s John Mueller has said a new date alone does nothing for search, and changing it for tiny edits can mislead readers (Search Engine Journal, 2021). The update has to be real.
• Traffic is not the same as fit. Some of HubSpot’s biggest posts covered topics far from its business, like resignation letters and famous quotes. Those were among the pages hit hardest when traffic dropped in late 2024.
• Search keeps moving. AI answers now show up before many blue links. Old posts still matter, but teams now have to think about whether AI tools can read and cite them, not just whether Google ranks them.
• Self-reported results. Most of the numbers come from HubSpot’s own blog, so treat them as a strong example, not a controlled study.

Sources

• The Blogging Tactic No One Is Talking About: Optimizing the Past, HubSpot Blog (Pamela Vaughan) (2025)
• Content lessons from the HubSpot blog team, David Meerman Scott (2014)
• HubSpot’s SEO collapse: What went wrong and why?, Search Engine Land (2025)
• Google’s John Mueller: Changing Dates on Pages Won’t Improve Rankings, Search Engine Journal (2021)
• HubSpot’s AI search experiments, Growth Unhinged (Aja Frost) (2026)

---

Use this case in: Analytics with GA4 · Content Strategy and Brand Storytelling · SEO and AI Discoverability · Content Strategy: What to Make and Why · All case studies

Brand names, images and trademarks belong to their owners and appear here for educational purposes only.