Most People Don’t Read

by

in

This is a qualitative study on a small number of anonymized users while collecting a very large number of datapoints from each one.

In December 2025, we published an article asking a simple question: Do you read or skim? We tracked 269 visitors using mouse movements, scroll patterns, and time-on-page data, then asked them to self-identify their reading behavior.

The goal was to compare self-reported behavior against actual behavior—and see how things have changed since our original 2015 study.

The Headline Number

2015
56%
self-identified as readers
2025
28%
self-identified as readers

In 2015, when we combined “read everything in full” (16%) with “read most, may skip some parts” (40%), we got 56% of respondents who considered themselves readers. In 2025, that number has dropped to just 27.7%.

That’s a 28-percentage-point decline in a decade.

The Numbers at a Glance

269
Total visitors tracked
52%
Completed the poll
72%
Self-identified as skimmers

How Long Do People Actually Stay?

We tracked how long visitors remained on the page before submitting their response. The retention curve is steep:

Percentage of visitors still on page at each time mark

After just 30 seconds, two-thirds of visitors had already left or submitted their response. Only 11% were still engaged after one minute.

22s
Median time on page
Using median to exclude outliers

The Honesty Problem

Here’s where it gets interesting. We compared what people said about their reading behavior against what they actually did.

We defined “reader behavior” as: spending more than 30 seconds on the page AND scrolling past 75% of the content. Under this definition:

Behaved as
Reader
Behaved as
Skimmer
Claimed “Reader”
16
17
Claimed “Skimmer”
30
62
48%
“Readers” who actually read
67%
“Skimmers” who actually skimmed

Only half of self-identified “readers” actually exhibited reading behavior. Meanwhile, 30 people who called themselves skimmers actually spent meaningful time with the content—perhaps they’re humble, or have simply accepted the cultural norm that “nobody reads anymore.”

The Aspirational Reader

We found 15 visitors who clicked “Reader” but spent less than 20 seconds on the page. We call these aspirational readers—people who believe they read, or want to believe they read, but don’t.

Engagement by Traffic Source

Where visitors came from significantly affected their reading behavior:

Self-identified reader percentage by traffic source
LinkedIn
31.4%
Twitter/X
16.7%
Facebook
11.1%

LinkedIn visitors were nearly 3x more likely to identify as readers compared to Facebook visitors. Twitter/X fell in the middle—perhaps unsurprising given the platform’s emphasis on rapid-fire content consumption.

Traffic Source Median Time
Twitter/X 27.3s
LinkedIn 25.5s
Facebook 17.2s

Mobile vs Desktop

Mobile visitors made up 70% of our traffic—reflecting the broader shift in how people consume content.

Self-identified reader percentage by device
Mobile
29.9%
Desktop
22.7%

Counterintuitively, mobile users were more likely to self-identify as readers. However, desktop users generated significantly more engagement signals—a median of 154 tracked events compared to just 59 on mobile. This likely reflects the richness of mouse movement data versus touch interactions.

The Engagement Score

We created an “engagement score” based on total tracked interactions: mouse movements, scroll events, and clicks. Here’s how self-identified readers compared to skimmers:

Metric Readers Skimmers
Time on page (median) 23.7s 21.0s
Total events (median) 101 82
Scroll events (median) 57 45
Scroll depth (median) 97.3% 92.9%

Self-identified readers showed 23% more total engagement and 28% more scroll events. But the time difference was minimal—just 2.7 seconds. Both groups reached near-complete scroll depth, suggesting most visitors at least scrolled through the entire article, even if they didn’t read it.

Key Findings

  • Reading has declined dramatically. Only 28% of visitors self-identified as readers in 2025, down from 56% in 2015—a 28-point drop.
  • Self-reporting is unreliable. Half of “readers” didn’t actually read. People overestimate their reading habits.
  • 30 seconds is the cliff. Two-thirds of visitors are gone within 30 seconds. Your first paragraph is everything.
  • Platform shapes behavior. LinkedIn visitors are 3x more likely to read than Facebook visitors. Know your audience.
  • Scrolling ≠ reading. Both readers and skimmers scrolled through 90%+ of the page. Scroll depth alone is a poor proxy for engagement.
  • Some skimmers are actually readers. 30 self-identified “skimmers” exhibited reading behavior. The social norm may be to downplay reading habits.
“We’ve shifted from a culture of reading to a culture of scanning. The question isn’t whether people will read your content—it’s whether they’ll give you 20 seconds to prove it’s worth reading.”

What This Means for Content Creators

If you’re creating content for the web, here’s the uncomfortable truth: most people won’t read what you write. Not because your content is bad, but because scanning has become the default mode of information consumption.

This doesn’t mean long-form content is dead. It means the first 20 seconds matter more than ever. Front-load your value. Make your key points scannable. And accept that the minority who do read will be your most valuable audience.

Methodology

This study tracked 269 unique visitors to a single article page on dejan.ai between December 25-30, 2025. We collected anonymous mouse movement, scroll, and click data using client-side JavaScript, and asked visitors to self-identify as “readers” or “skimmers” via an embedded poll.

All statistics use medians rather than means to account for outliers (visitors who left browser tabs open). Behavioral classification used thresholds of >30 seconds time-on-page AND >75% scroll depth to define “reader behavior.”

Traffic sources: LinkedIn (34%), Twitter/X (20%), Direct (25%), Facebook (7%), Other (14%).


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *