What is content value, and why isn’t it a pageview?
Content value is how much a piece of content matters to the people who use it. Here is what it means, why traffic alone can't measure it, and how to start measuring it properly.
What content value means
Content value is how much a piece of content matters to the people who use it. A guide that someone reads to the end and acts on is worth more than an article that a hundred people open and abandon in three seconds, even though the second one shows up as more traffic.
That sounds obvious, and yet the web mostly measures something else. It measures arrival: a visit, a click, a pageview. Value shows up in what happens after arrival: whether the content was read, whether it answered the question, whether the person trusted it enough to stay, act or return.
Why pageviews are incomplete
A pageview is recorded when a page loads. Nothing about it depends on whether the visitor read a sentence. That makes it easy to count, easy to compare and easy to sell, which is why it became the default unit of content success.
It also leaves out almost everything you'd want to know:
Attention. Did the visitor scroll, read, or bounce immediately?
Depth. Did they reach the part of the page that mattered?
Usefulness. Did the page do its job, whether that was answering a question, teaching a skill or supporting a decision?
Return. Did they come back, or tell someone else?
Intent. Were they the right audience, or a stray click from a misleading headline?
Pageviews can also be inflated by things unrelated to value: a sensational headline, a lucky share, a redirect, or automated traffic. When a number can go up without the content getting any better, it can't be the whole definition of quality. There's more on this in why pageviews are a poor measure of content success.
Engagement versus traffic
Traffic measures how many people arrived and where they came from. Engagement measures what they did with the content once they were there. Both are useful, and they answer different questions.
Traffic answers: is anyone finding this? Useful for distribution, SEO and campaigns.
Engagement answers: is it worth finding? Useful for editorial decisions, product decisions and, eventually, pricing.
A page can have high traffic and low engagement (a clickbait headline), or low traffic and very high engagement (a niche guide that exactly serves a small audience). Only one of the two tells you the second page is worth a lot to the people who found it.
What makes content useful
Engagement is a signal of usefulness, not the thing itself. Useful content tends to share a few properties, and it helps to keep them in mind so you don't confuse a signal with the goal:
It answers the question the reader actually came with, early and clearly.
It has depth the reader can't get from a summary: original reporting, first-hand experience, real data, a clear argument.
It is trustworthy: sourced, specific, and honest about what it doesn't know.
It is well structured, so a reader can scan it or read it end to end.
It stays useful: evergreen guides can earn attention for years.
How to measure content quality and performance
No single metric captures value. A practical approach is to look at several signals together and treat each one with suspicion:
Scroll depth: how far down the page visitors get. A good proxy for whether they got past the intro. It doesn't prove they read every line.
Time on page: a rough guide to attention, easily distorted by a tab left open.
Return visits: a strong sign that a source is trusted, but the hardest to measure without tracking individuals.
Outcomes: subscriptions, signups, purchases, shares, replies. The clearest evidence, but usually rare and lagging.
Qualitative signals: emails from readers, citations, the questions people ask afterward.
If pageviews are incomplete, what should replace them? Not a single replacement, but a shift in the unit of account: from arrivals to engaged visits, and from sites to pages.
Engaged sessions: count visits that cleared a meaningful bar (for example, scrolled well into the page) rather than every load.
Per-page engagement: compare pages against each other, so a small, deeply read page isn't buried by a big, skimmed one.
Cohort comparison: judge a page against pages with similar traffic, not against the whole site.
Outcome tracking: tie pages to the actions they lead to, where that's appropriate and privacy-safe.
Where Widgetail fits
Widgetail is building a measurement and economic layer around real engagement with web content: measure it, score it, and eventually let that value become an economic signal. Here is how far along each part is.
Measuring engagement
Done
Pageviews, unique sessions and average scroll depth from one script tag
Traffic over time, top pages, and change against the previous period
Anonymous sessions: no IP addresses, no fingerprints
Building now
Reading behavior: engaged time, bounce rate, pages per session and a scroll-depth funnel
Audience breakdowns: traffic sources, devices, browsers, countries and busiest hours. The script already sees these; the Console doesn't report them yet
Planned
Tracking controls: honor Do Not Track, exclude paths, strip query strings
Data retention settings and CSV export
A privacy-safe way to see return visits, if one can be designed without identifying people
Scoring content value
Each page gets a score from how people read it. Here is what that looks like, using illustrative pages:
Content valueBetaIllustrative data
Sample content value scores with the four signals behind each score
Page
Score
Engagement
Depth
Repeat
Interactions
A field guide to reducing cloud costs/guides/reducing-cloud-costs
82High
86
81
58
92
Why we moved off ads/essays/why-we-moved-off-ads
85High
90
84
61
94
The 2026 buyer's guide/reviews/2026-buyers-guide
75High
79
75
47
85
Getting started/docs/getting-started
69Medium
70
68
72
66
Pricing/pricing
52Medium
58
55
50
45
About us/about
31Low
40
42
28
15
Each page is scored 0–100 from four signals, compared against pages with similar traffic.
Done
A 0–100 score for every page, built from four signals: time spent reading, scroll depth, repeat views and interactions
Scores compared within pages of similar traffic, so a small, deeply read page isn't buried by a big, skimmed one
A low-sample flag when a page has too little traffic to trust its score
A Console view of every score and its signal breakdown, tested on synthetic data
Building now
Validating scores against real site traffic
Adding the time and interaction signals the score draws on
Planned
Score trends against the previous run
Safeguards against artificially inflated engagement
Cross-site content authority signals, built with strict privacy thresholds
Each page gets a 0–100 score from four signals: time spent reading, scroll depth, repeat views and interactions. Pages are compared with pages of similar traffic, so a small, deeply read page isn't buried by a big, skimmed one, and pages with too little traffic are flagged as low-sample. The blend is a beta and can change as it is validated against real traffic.
Is engagement the same as content quality?+
No. Engagement is a signal that content may be useful, not proof. Sensational content can be engaging without being good, and a brief, perfect answer can be valuable with very little engagement. That's why value measurement should combine several signals and be read with judgment.
What is the best alternative to pageviews?+
There isn't a single replacement. A better approach is to count engaged visits instead of every page load, compare pages against similar pages, and connect content to outcomes where you can. Scroll depth is one practical, privacy-friendly signal to start with.
Can you measure content value without tracking individuals?+
Yes, for many signals. Scroll depth, session counts and per-page engagement can be measured with anonymous sessions and no IP addresses or fingerprints. What you give up is cross-visit tracking of the same person, which is why some signals, like long-term return rate, are harder to measure privately.