A data essay on how we consume stories

The
Shrinking
Story

Are we losing patience — or just becoming honest about our limits?

Scroll to read
01 — The Confession

150 minutes of TV.
A 2-hour film we couldn't start.

We've been meaning to watch a film. A good one. Two hours. Everyone recommends it. But every time we open the app, we scroll past it and choose a series instead — the episodes are only thirty minutes.

I can't commit to a two-hour film. But I'll watch five episodes back to back without noticing the time.

— A confession most of us recognise

Five thirty-minute episodes is 150 minutes. The film is 120. We don't have short attention spans. We have an aversion to a particular kind of upfront promise.

It's a change in the unit of commitment — and it's been happening across every storytelling medium for well over a century.

150 minutes actually watched
5 × 30-min episodes
vs
120 minutes of the film
we "couldn't commit to"
· · ·
02 — The Museum of Attention

The unit of story, by era

It's been happening for over a century. From Tolstoy to TikTok — the typical unit of story consumed by each generation. Hover to read more.

← → Drag to explore · Hover to expand
03 — The Compression

Average content unit, medium by medium

The same human appetite for story — in a shrinking container.

The TikTok number looks like collapse. But the average viewer watches 95 minutes of TikTok per day — roughly 271 videos. 271 stories. They're not experiencing less narrative. They've simply removed the moment of commitment entirely.

James Patterson understood this long before algorithms did. His chapters are 800 words. Each one ends mid-breath. Dickens did the same thing in 1836, one serialized installment at a time.

The chapter cliffhanger is not new.
Dickens invented it in 1836.

04 — The Paradox

But here's what the unit data doesn't show

Active daily story consumption per person (US adults) — background TV and radio excluded. The unit shrinks; the total grows.

Active vs passive: Nielsen distinguishes "intentional" from "background" viewing — a TV on while cooking is not counted here. Radio is excluded entirely. What remains is time where a person is the primary audience of a story: watching, reading, or scrolling with intent. This methodology follows Nielsen's Total Audience Report (2010–2024) and Datareportal Digital 2024.
Linear TV (active)
Streaming video
Social video
Books + Cinema

Estimated from Nielsen Total Audience Reports, Datareportal Digital 2024, Pew Research Center. US adults. Figures are approximations — precise breakdowns vary by methodology and year.

In 2005, the average US adult spent roughly 3.6 hours a day actively watching or reading stories. By 2024 that figure is closer to 6.2 hours — a 70% increase.

Linear TV shed two hours a day. Streaming and social video more than made up the difference. The total keeps rising. Each new format lowers the barrier to starting — and so we start more often.

We watch more. We read more.
We just commit less per sitting.

05 — The Literary Record

One medium didn't compress

Man Booker Prize winners, 1969–2023. While every other format surrendered to the shrink, the literary novel held its ground — and occasionally pushed back.

Each winner
9-year moving average
90,000 words (typical novel)

* Word counts approximate. Sources: publisher data, academic analyses, fan-sourced databases.

Midnight's Children (1981): 210,000 words. The Luminaries (2013): 362,000. But alongside these monuments — Julian Barnes won in 2011 with The Sense of an Ending at just 36,000 words. Ian McEwan's Amsterdam (1998): 38,000.

The judges kept choosing brevity, then length, then brevity again. The craving for depth never left. It just moved off-screen.

The long novel didn't disappear. It became an act of defiance.

06 — The Last Compression

Even when we commit,
we speed up.

The craving for long-form is real. But patience for it is another matter. Audiobook revenue has grown every year for a decade — long-form stories reclaimed in commute time. Look at how people actually listen.

~64% listen at 1.25× speed or faster
~28% listen at 1.5× or above

A 12-hour audiobook at 1.5× becomes 8 hours. We reach for something long and slow — then ask it to go faster.

07 — The Reframe
Long stories aren't
disappearing.
Our way of
consuming them
is mutating.

The Victorian reader who bought the next Dickens installment every week was doing exactly the same thing. So was the person who stayed up until 3am because Patterson's chapters were too short to stop at.

The unit gets smaller. The appetite stays the same.

Data notes
Man Booker Prize word counts are approximate estimates. Bar chart figures represent estimated average content duration per unit in each medium's dominant era. Consumption chart sourced from Nielsen Total Audience Reports (2010–2024), Datareportal Digital 2024, and Pew Research Center; figures represent US adults, active/intentional consumption only — background TV and radio are excluded. Audiobook speed data from Audible/Spotify listener surveys. Daily TikTok usage from ByteDance and independent research (2023–24). All figures are approximations.

Further reading
James Patterson on chapter length · Naomi Baron, "Words Onscreen" · The Pudding, various · Nicholas Carr, "The Shallows"