Somewhere between page one and page forty-seven, you lose them. That’s not a gut feeling — it’s a dataset. In Q1 2026, Jellybooks completed its largest reader-engagement study to date, tracking 112,000 readers across 340 titles released between January 2025 and March 2026. The headline number: 61% of DNFs happen before the end of chapter three. The secondary number, the one that’s currently making acquisitions editors at every major imprint genuinely nervous: among readers aged 18–34, that figure climbs to 74%.
“We’ve been sitting on this data for six weeks trying to figure out how to present it without triggering a panic,” says Andrew Rhomberg, Jellybooks’ CEO, who shared an early cut of the findings with BookTok Times ahead of the full June 18 release. “The short version is: readers are not patient. The long version is: readers have never been patient, but now they have the language to say so publicly, and BookTok gave them that language.”
The industry is calling it the DNF Crisis. Publishers are calling it a narrative pacing emergency. And the creators — the @perpetualpages and @fantasybookjunkies of the world, with their combined 40-plus million followers — are calling it exactly what it is: readers finally admitting what they’ve always done.
What does the 2026 reader data actually tell us about why people quit books?
The Jellybooks study, cross-referenced with Goodreads’ internal review-sentiment analysis (shared selectively with a handful of Big 5 editors under NDA, but leaked in broad strokes to BookTok Times by two separate sources), isolates four primary abandonment triggers:
- Passive protagonist in chapter one — cited by 68% of DNF respondents as the single largest factor
- Worldbuilding front-loaded before emotional stakes are established — 54%
- No clear point-of-view interiority in the first 1,000 words — 49%
- The lead character makes no consequential decision before page 30 — 41%
“The passive protagonist number should end careers,” says Emma Noyes, editorial director at Wednesday Books, who was not involved in the Jellybooks study but has been tracking its early data through industry channels. “I’ve said it in acquisition meetings. I’ve said it in revision letters. And I’m still getting manuscripts where the heroine wakes up, looks in a mirror, and goes to breakfast. That manuscript is dead on arrival in 2026.”
Why are female fantasy readers gravitating toward darker, more morally complex male leads?
The data gets more granular — and more commercially consequential — when Jellybooks breaks it out by genre and gender. Among women reading fantasy in 2026, the highest completion rates belong to books featuring what the study clinically categorizes as “antagonist-adjacent male leads” — characters the researcher team describes as operating in sustained moral grey zones, with explicit capacity for violence and clearly articulated inner conflict.
Completion rates for books with this character architecture: 83%. Completion rates for books featuring conventionally heroic male leads in the same genre: 61%.
“Readers are not looking for a good man. They are looking for a dangerous man who chooses to be good specifically for the female lead. That specificity is the entire emotional engine.” — Cait Corrain, author of Crown of Starlight, speaking at the London Book Fair BookTok Summit in April 2026
This tracks with what Macmillan’s data science team has reportedly been calling internally “the Cardan Coefficient” — a reference to the Folk of the Air lead whose cultural footprint continues to distort acquisition heuristics three years after the series’ peak. The coefficient attempts to quantify the possessive-protector-with-genuine-menace quotient in submitted manuscripts. Two acquisition editors at Macmillan confirmed the internal metric exists. Neither would go on record.
For BookTok creators, the darker-male-lead preference isn’t news — it’s content strategy. “Every single one of my top-performing videos in 2026 has been about a male lead who does something wrong and means it,” says @darkpagesafter, who crossed 3.1 million followers in May. “The second I put up a wholesome love interest, my retention drops. The algorithm knows what the algorithm knows.”
Does emotional pain actually outperform happy romance in sales and completion data?
The short answer is yes, and the margin is widening. Goodreads’ internal review-sentiment data, which the company shared in summary form at its Publisher Partner Summit in March 2026, shows that books tagged with sentiment markers like “destroyed me,” “I sobbed,” and “emotionally wrecked” carry a 34% higher reread rate and a 27% higher gifting rate than books tagged with “sweet,” “wholesome,” or “feel-good.”
The gifting rate is the number that should be tattooed on every marketing VP’s forearm. Gifting drives secondary sales without requiring a single additional marketing dollar. When a reader is emotionally wrecked by a book, they make it someone else’s problem. That’s a distribution channel.
“Pain is a referral engine. Joy is a private experience. We’ve known this intuitively forever. Now we have the numbers.” — Sandhya Menon, author and BookTok-native panelist, speaking at BookExpo Spring 2026
The audiobook data compounds this. Libro.fm’s listener analytics, shared in a limited partnership report with select agents in May 2026, show that listeners of “emotionally devastating” audiobooks — defined by the platform as titles receiving three or more five-star reviews using grief-related language — have an average relistening rate of 2.3x per title. For comparison, listeners of romantic comedies relistenat 1.1x.
“The relisten number tells you something profound about how readers use these books,” says Deb Shapiro, head of audio acquisitions at Macmillan Audio. “They’re not just consuming a story. They’re using the emotional experience as a resource. That’s a fundamentally different relationship with a book than we’ve historically planned for.”
How are Gen Z and millennial readers diverging in their expectations in 2026?
The generational split in reader expectations has sharpened to the point where some acquisition editors are now building separate targeting briefs for books they believe will skew differently by age cohort. The Jellybooks data supports the instinct.
Gen Z readers (18–27) in the study prioritize, in order:
- Immediate emotional stakes — the character must want something on page one
- Morally complex relationships rather than aspirationally pure ones
- Ambiguous or bittersweet endings over clean resolutions
- Dark themes treated without apology or authorial editorializing
Millennial readers (28–42) prioritize:
- Emotional safety within darkness — darkness with a guaranteed emotional payoff
- Slow-burn tension with high reward at culmination
- Strong internal character monologue and psychological interiority
- Clear, earned emotional resolution, even within dark narratives
“Gen Z wants the wound. Millennials want the wound and the bandage,” says Rhomberg. “The smartest authors in 2026 are the ones writing books that feel like they won’t give you the bandage — and then give it to you in a way that makes you feel like you didn’t need it but are grateful anyway.”
This divergence is creating real tension in how BookTok creators pitch books to their audiences. Creators whose followings skew Gen Z — including @pageswithpurpose (4.8M followers, median follower age 22) — are increasingly championing books with unresolved or ambiguous endings that would have been commercial death sentences as recently as 2022. Creators with millennial-heavy audiences are still defaulting to the emotional-payoff promise in their hook language.
Why do readers reread dark books more than any other category?
The reread economy — estimated at $340M in secondary sales impact annually, based on BookTok Times’ own analysis of Goodreads shelf-movement data — is almost entirely driven by dark romance and dark fantasy. The Jellybooks study attempts to explain the psychology, and its conclusions are striking.
Readers who reread dark books report, in exit surveys, that they are primarily seeking one of three things: emotional regulation (the book provides a controlled environment to experience intense feeling), character companionship (they miss the specific interiority of a character they found themselves inside), or anticipatory pleasure (they are re-experiencing the buildup to a moment they already know is coming, which they describe as uniquely satisfying).
“The reader already knows the villain will choose her. The second read is about savoring the inevitability. That’s not nostalgia — that’s an entirely different emotional technology.” — Jennifer L. Armentrout, whose backlist reread rates are among the highest tracked in the Jellybooks dataset, in a BookTok Live conversation in May 2026
For publishers, the reread signal is increasingly being used as a predictive acquisition metric. If early ARC readers are flagging that they “need to immediately reread,” that signal now carries more weight in some imprints’ greenlight processes than aggregate star ratings. “A book with eighty percent five-star reviews and no reread mentions scares me a little,” says Noyes. “It means it was satisfying once. That’s fine. But the books that change your life make you read them twice.”
What does all of this mean for acquisitions and editorial strategy going into 2027?
The Jellybooks data, combined with Goodreads’ sentiment analysis and Libro.fm’s listener metrics, is assembling into a coherent mandate that the industry is only beginning to institutionalize. The reader of 2026 — across age cohorts, across formats — is demanding emotional immediacy, moral complexity, and the particular pleasure of pain that they trust will mean something.
The slow opening is a liability. The safe male lead is a liability. The clean, unearned happy ending is a liability. These are not aesthetic preferences. They are commercial variables with measurable dollar values attached, and the data is now granular enough that editors who ignore them are making a documented choice to leave revenue on the table.
“Publishing spent a decade trying to figure out what BookTok wanted,” says Rhomberg. “The answer was always the same. Readers want to feel something real inside something impossible. The genre doesn’t matter. The length doesn’t matter. The format doesn’t matter. Give them a character in genuine pain making genuine choices, and they will not put the book down. Give them anything less, and they’ll be on TikTok within six minutes telling three million people why they quit.”
The full Jellybooks 2026 Reader Engagement Report publishes June 18. Advance copies are available to verified publishing industry contacts through the Jellybooks partner portal.