- Pages tagged bottom-of-funnel (BOFU) often show up early in a buyer's path, and top-of-funnel (TOFU) pages often show up right before a deal
- Glossary pages appear across every stage, which most teams don't expect
- A single page view can sit for years before the account becomes a qualified opportunity
- Looking at the pages that influenced a deal tells you more than looking at the page that started it
Most content plans start with a funnel diagram. Blog posts sit at the top, comparison pages and case studies sit at the bottom, and everything else falls in between. It's a tidy way to plan, and it's easy to explain to a leadership team.
The trouble starts when you compare that diagram to what buyers actually do. When we pull page-view history for people tied to real deals, the picture is a lot less orderly.
In this post, we'll walk through what that data tends to show, why it happens, and what to change about how you plan and measure content.
What deal-level content data looks like
One of the most useful things we do with customers is build a detailed report from their HubSpot data (or whatever CRM and analytics they have). We want to know:
- The order in which people view pages on the site
- Which pages are connected to deals and which only to leads
- How those pages roll up to topics
- What type each page is
- When each visit happened
We've now done this dozens of times. The overall takeaway is the same every time: it's incredibly messy.
Your TOFU and BOFU labels are probably wrong
Content you thought of as bottom-of-funnel shows up as someone's first visit. Content you thought of as top-of-funnel shows up in the sessions right before a deal opens.
That doesn't mean funnel stages are useless as a planning tool. They're a reasonable guess about intent. But a guess is all they are, and the data regularly disagrees with it.
A few reasons this happens:
- Buying committees. In B2B, several people research independently. One person reads a comparison page on day one while another reads an explainer a week before the deal opens.
- Search and AI answers send people to odd places. Someone with a mid-funnel question may land on whatever page answers it best, whatever label you gave that page.
- People return for reference. Buyers late in the process go back to explainers to check a definition or share it with a colleague.
Glossary pages show up everywhere
The biggest surprise has been glossary pages. We wouldn't have guessed it, but they're spread across every stage of the journey. They appear in first visits, in the middle of long sessions, and in the final stretch before a deal.
It makes sense in hindsight. A glossary page answers a narrow, specific question, and buyers have those questions at every stage. A newcomer wants to learn the term. A buyer evaluating vendors wants to confirm they understand it. Someone writing an internal memo wants a clean definition to paste in.
If you've been treating glossary content as low-value SEO filler, check the deal-level data before you cut it.
Journeys can run for years
The data also shows how long things take. While reviewing one customer's deals, we found a page view from five years earlier. Then nothing, and then the account became a qualified opportunity. We don't know whether the deal closed, but it became a real opportunity.
Five years is on the long end, though it isn't as unusual as you might expect. These weren't small purchases, either. They were deals worth $1 million or more, and buying cycles that big can take a while.
Someone read a page, took a different job, spent years elsewhere, and eventually landed in a buying cycle. When they did, the brand was already familiar. That's content doing its job, even though no dashboard would have credited it.
The lesson we keep relearning is that you have very little control over the real customer journey. Most of the time, you're making your best guess, and it can take a long time to pay off.
How to look at your own data
You don't need a custom analysis to start. A simpler version looks like this:
- Pick a stage to examine. Qualified leads, qualified opportunities, and deals are all useful. We'd stay cautious about going all the way to closed-won, since sales execution affects that outcome as much as content does.
- Pull the page history for people tied to those records. HubSpot works well for this, and a connected CRM and analytics setup makes it much easier.
- Include other people at the same account. Often, the person who read your content isn't the person attached to the deal.
- Focus on views before the deal was created. Views after creation are fair to look at, but pre-deal views are usually more convincing.
- Group pages by topic and page type, then compare that to the stage you'd assigned.
A few gotchas will make your numbers look smaller than reality. Product sign-ups often come into HubSpot as offline sources, with no page history attached. Tracking cookies are also missing a lot of the time. In our experience, you might see page history for only about 40–60% of visitors. Treat your results as a floor.
What to change
If your funnel is messier than your plan assumes, a few practical shifts follow:
- Stop judging a page only by where it sits in the funnel. Judge it by whether it shows up in the history of real deals.
- Be patient with long-lag content. A page may not show an effect for months or years.
- Keep the quiet workhorses. Glossary pages and similar reference content may be doing more than your traffic reports suggest.
- Look at influence, not just origin. Ask which pages influenced the people on a deal, instead of only which page started the relationship.
Conclusion
Funnel labels are a useful way to plan, but they don't describe how buyers move. Deal-level data shows BOFU pages at the start, TOFU pages near the end, glossary pages everywhere, and journeys that can span years.
The fix isn't a better diagram. It's looking at the pages real buyers viewed and letting that shape what you publish next. At Ercule, we've found this approach tells you both whether your content program is valuable, which it almost always is, and which specific tactics deserve more investment.
Frequently asked questions
Does this mean funnel stages are useless?
No. They're a reasonable starting guess about buyer intent and help with planning. Just check them against deal-level data instead of assuming they hold.
Why do glossary pages appear at every stage?
They answer narrow questions, and buyers have narrow questions throughout the process. Newcomers learn terms, evaluators confirm them, and champions borrow definitions for internal documents.
How do I find which pages influenced a deal?
Pull the page-view history for contacts associated with the deal and, ideally, other contacts at the same account. HubSpot, GA4, PostHog, and Amplitude can all provide versions of this.
Should I count page views after the deal was created?
You can look at both periods. Views before the deal was created are usually more convincing, so most teams focus on those.
Why does my data show so little page history?
Common causes are product sign-ups that arrive as offline sources, tracking pixels that don't fire consistently, and cookie blockers. In our experience, cookie-based history may cover only 40–60% of visitors.
Should I measure all the way to closed-won?
It's tempting, but closed-won depends heavily on the sales team. Qualified leads, qualified opportunities, and deals in progress are cleaner stages for judging content.
Is a five-year journey normal?
It's a record in the data we've seen, but long gaps aren't rare, especially for large B2B deals with buying committees.
