By: You know people read your document. The analytics say so. Average time: 6 minutes. 47 views this week. Looks great in a report.
But which parts did they actually care about?
Did they spend those 6 minutes on your case studies or your pricing page? Did they read the introduction carefully and skip the methodology? Did everyone bail at page 8 and you just don't know it because the average time still looks acceptable?
Overall metrics tell you the document is being consumed. Page-level analytics tell you what's working inside it.
Why Page-Level Data Matters
Think about it this way. You send a 12-page proposal. The overall engagement looks healthy. But page-level data reveals this:
- Pages 1-2 (Introduction): 90% of viewers, 40 seconds each
- Pages 3-4 (Our Approach): 55% of viewers, 25 seconds each
- Pages 5-6 (Case Studies): 78% of viewers, 65 seconds each
- Pages 7-8 (Solution Details): 48% of viewers, 30 seconds each
- Pages 9-10 (Pricing): 88% of viewers, 70 seconds each
- Pages 11-12 (Team & Next Steps): 32% of viewers, 15 seconds each
Now you know things. People open the introduction and jump to case studies and pricing. They skip your approach section and mostly ignore team bios. Your solution details page is losing half your readers.
Without this data, the document looks fine. With it, you know exactly what to fix. Move case studies and pricing earlier. Condense or cut the approach section. Either strengthen the solution details page or fold the key points into the introduction.
Analytics expert Avinash Kaushik has been pushing this principle for years: stop looking at aggregate numbers and start looking at where the engagement actually happens. "Dump the vanity metrics," he wrote. Aggregate read time is the vanity metric of document analytics. Page-level data is where the real story lives.
How to Read a Page Engagement Report
Most page-level reports show you two things per page: how many people viewed it and how long they spent. Both matter, but they tell different stories.
High Views, High Time
The page is interesting and people are reading it carefully. This is your strongest content. For a sales document, this is usually pricing or ROI sections. For marketing content, it's often the sections with concrete examples or data.
Whatever's on these pages, do more of it.
High Views, Low Time
People land on the page but leave quickly. This happens with introductions and tables of contents. Readers are scanning, not reading. That's normal for navigational pages.
But if a content-heavy page gets high views and low time, something is wrong. The content might be too dense. Too much text with no visual breaks. Or the headline promised something the body didn't deliver.
Low Views, High Time
A hidden gem. The people who find this page love it, but most readers never get there. This is common for pages buried deep in a long document. The fix is usually structural: move this content earlier, or create a shortcut to it from the introduction.
Low Views, Low Time
Nobody visits this page, and those who do leave immediately. This is the content equivalent of a dead zone. If removing it wouldn't lose anything important, remove it. If the information matters, it needs to be rewritten, redesigned, or relocated.
What the Patterns Reveal
Individual pages are useful. Patterns across many readers are where real insights emerge.
The Jump Pattern
Readers view page 1, then skip to page 8. This tells you people are looking for something specific. If page 8 is pricing, they want the bottom line. If page 8 is the conclusion, they want the summary without the buildup.
When you see this consistently, your document is probably too front-loaded with context people don't need. Get to the point faster.
The Drop-Off Cliff
Engagement is strong through page 5, then falls off a cliff at page 6. Something on page 6 kills momentum. Go look at it. Common culprits: a wall of text with no headings, a topic shift that feels disconnected, or a section that repeats what was already covered.
A drop-off cliff usually means you stopped respecting the reader's time at a specific point. Find that point and fix it.
The U-Shape
High engagement at the beginning and end, low in the middle. Classic structure problem. People read the intro, skip to the conclusion, and ignore the middle. Your middle sections need to earn attention, not just fill pages.
The Scatter
No clear pattern. Some people read pages 2, 5, and 9. Others read 1, 3, and 7. Everyone has a different path. This usually means the document serves multiple audiences and each one is looking for something different. Consider splitting it into shorter, targeted pieces instead of one long document.
How to Improve Weak Pages
You've found the pages where readers bail. Now what?
Shorten Them
The most common fix is the simplest one. Long pages lose people. If a page has 500 words and the one before it had 150, the reader hits a wall. Break it up. Cut the filler. Get the key point across in fewer words.
Add Visual Breaks
A page of solid text is intimidating. Headers, bullet points, images, pull quotes, diagrams, bold key phrases. Anything that breaks the visual monotony and gives the eye places to land. People scan before they read. Give them something to scan.
Move the Good Stuff Up
If your best content is on page 10, most readers will never see it. They're gone by page 6. Move the high-value content earlier. Your pricing breakdown, your key results, your main differentiator. Whatever readers actually want should be within the first few pages.
Cut the Dead Weight
Some pages don't need to exist. Company history? An appendix of team bios? Methodology details? If analytics consistently show zero engagement on these pages, cut them. Or move them to a separate document that interested readers can request.
The goal isn't to make every page equally engaging. It's to make sure no page actively drives people away.
What's a Good Sample Size?
Page-level data is only useful if enough people have viewed the document. With 5 readers, one person's unusual behavior skews everything. With 50, you have a pattern.
For general trends (which pages are weakest, where drop-off happens), 20-30 viewers is a reasonable minimum. For finer-grained decisions like reordering specific sections, aim for 50 or more.
If you don't have enough views yet, wait. Or combine data across similar documents. If every sales proposal you send shows the same drop-off at the methodology section, that's a pattern even if each individual proposal only has 10 viewers.
Putting It Into Practice
Start with your most-shared document. The proposal template your sales team uses every week. The product guide marketing sends to every lead. The training manual every new hire receives.
Pull up the page-level analytics. Look for the patterns described above. Find the weakest page. Fix it. Then check back in a month and see if completion rates improved.
Small changes compound. Moving your pricing section from page 10 to page 4 might increase the percentage of readers who see it from 48% to 85%. Cutting a two-page methodology section down to a half-page summary might reduce your drop-off rate at that point by 30%.
You won't know until you look. And once you look, the changes are usually obvious.
Related reading:
FAQs
How many views do I need before page-level data is reliable?
At least 20-30 viewers gives you a reasonable pattern. Below that, individual reading habits can skew the data. For high-stakes decisions like restructuring a sales proposal, aim for 50+ views.
Is page time the same as page importance?
Not always. A page with high time might be genuinely interesting, or it might be confusing. Look at time in combination with completion rate and whether readers continue to the next page or bail.
Does page-level tracking affect document load time?
No. The tracking happens in the background as readers navigate pages. There's no noticeable impact on how fast the document loads or how smoothly pages turn.
Can I see which pages a specific person viewed?
Yes. If the reader accessed through a tracked link or entered their email, you can see their individual page-by-page journey — which pages they viewed, how long on each, and in what order.
Does this work for privacy-conscious audiences?
Page-level analytics track engagement with your hosted content, similar to how any website tracks which pages visitors view. No personal data is collected unless the reader voluntarily provides it through an email gate. This falls under standard web analytics practices.