Sharan Initiatives
๐Ÿ“š
๐Ÿ“šLiterature

Running an Online Book Club That Doesn't Quietly Die

Most online reading groups collapse in four months for the same handful of reasons. What actually keeps a discussion alive when nobody is in the same room.

By Taresh Sharan ยท PhD, IIT BHUโ€ขJanuary 30, 2026โ€ข8 min read

Almost every online book club I have been part of has died the same way. Strong start, nine enthusiastic people, a genuinely good first discussion. Then a month where three people did not finish. Then a month where the organiser is the only one posting. Then nothing, and the group chat sits there for a year until someone leaves it.

This is so consistent that it is worth treating as a design problem rather than a failure of individual commitment.

It is also worth starting with a corrective on scale. Book clubs are a small phenomenon, not a mass movement: Pew Research Center has found that only a small single-digit percentage of US adults had taken part in a book club in the previous year. If it feels like everyone is in one, that is because the people in them talk about it. This matters practically: you are not tapping into a large latent demand, and you should size your expectations accordingly.

Why they die

The book was too long, too often. The single most common killer. A four-hundred-page literary novel every month is roughly ten hours of committed reading, on a deadline, alongside everyone's actual life. Miss one month and you cannot participate; miss participation once and you have no reason to come back. The attrition is arithmetic.

Synchronous meetings across time zones. A video call that works for everyone is a call that works badly for several people, and those are the people who drift.

Nobody owns it. Groups without a specific person responsible for choosing the book, posting the questions, and starting the discussion do not simply run at a lower level. They stop entirely, because the moment where someone should do something passes and nobody does.

Discussion by unstructured chat. A channel where people post reactions as they read produces spoilers, fragmentation, and a wall of text nobody can engage with after the fact. Reading at different speeds is the normal condition of a book club and the format has to accommodate it.

Nobody says they did not finish. The silence spiral is real. One person quietly did not read it, assumes they are the only one, says nothing, does not show up. Usually three others were in the same position.

What actually works

Shorter books than you think. Two hundred and fifty pages is a much better default than four hundred. You can always pick something long occasionally, with more time attached โ€” six weeks for a big novel rather than four. Completion rate is the metric that determines whether the group survives, and book length is the variable you control most directly.

Asynchronous by default, synchronous as a bonus. A written thread that opens on a fixed date, with a live call for whoever can make it, is far more robust than a call-only format. Async means the person in a different time zone or a busy fortnight still participates. It also produces better thinking โ€” people write more considered things than they say.

A named person per book. Not a permanent leader; a rotating one. Whoever chose the book is responsible for opening the discussion with four or five real questions and for replying to people during the first week. Rotating the role spreads the work and, more importantly, spreads the sense of ownership.

Structure the thread by section, not by chronology. Separate threads or clearly marked sections for the first third, second third, and end. This solves spoilers and lets slower readers join meaningfully rather than scrolling past discussion of an ending they have not reached.

An explicit norm that not finishing is allowed. Say it out loud, repeatedly. "Come anyway, tell us where you stopped and why" produces far more interesting discussion than silence, and a book that several people abandoned is often the best conversation the group will have.

Questions that produce discussion

Most book club questions are comprehension checks in disguise, and they get comprehension-check answers. What did you think of the ending. Did you like the protagonist. These close a topic rather than opening one.

The ones that reliably generate conversation share a property: they require a position that a reasonable person could disagree with.

Ask which character the author is least fair to. Ask what the book would lose if it were told by someone else. Ask where the writer lost their nerve. Ask whether the ending is earned or arranged. Ask what this book is arguing for, and whether you believe it. Ask about a specific paragraph โ€” reading a passage aloud and asking why it is built that way will do more for a group's reading than any amount of thematic discussion.

The other reliable technique is to ask people to bring something rather than answer something: one sentence they marked, and why. It gives everyone a contribution that requires no cleverness, and the accumulated passages tell you what the group actually noticed.

Choosing books without arguments

Rotation is the least contentious system: each member picks in turn, everyone reads it, no voting. It guarantees variety, it distributes responsibility, and it means every member occasionally reads something outside their preference โ€” which is most of the value of a book club over reading alone.

Pure majority voting tends to converge on a narrow band of taste, because the median preference wins every time and the median preference does not move. If you do vote, vote among a slate nominated by one person, so the slate carries someone's actual taste rather than the group's average.

One genuinely useful constraint: no book longer than some agreed limit without a group decision to extend the schedule. Put it in writing at the start, when nobody has a particular book in mind and the rule is therefore uncontroversial.

Size, and the silent members

Six to ten members is the working range. Below six, two absences kill a discussion. Above about twelve, most people stop feeling individually responsible for showing up, and a written thread becomes something to scroll rather than to join.

Expect that a third of any group will be readers rather than posters. That is not failure โ€” they are getting something out of it โ€” but if the ratio inverts the thing is over. The counter is to ask people directly, by name, occasionally: what did you make of the second part. It works, and it is slightly awkward, and it is the main job of whoever is running that month.

The variant worth knowing about

If every attempt at a discussion group has failed for you, try the silent book club format instead: people meet, at a set time, and read their own books in company for an hour, with a short conversation before or after about what they are reading.

It has none of the coordination problems. No shared book, no deadline, no possibility of not having finished. What it provides is the two things a book club actually delivers for most people โ€” a protected hour and the mild social pressure of other people reading โ€” without the machinery that usually breaks. Online, this works better than it sounds: a video call with cameras on and microphones off is a surprisingly effective place to read for an hour.

Why bother

The specific value of reading a book alongside other people is that you find out how differently it was read. You will be certain a character was sympathetic and discover that three people found them insufferable, and both readings will be defensible from the text. That experience โ€” not the accountability, not the socialising โ€” is what a book club offers that reading alone does not.

Everything above is in service of keeping the group alive long enough for that to keep happening. Which mostly means: shorter books, written discussion, somebody responsible each month, and permission to turn up having read half.

Tags

LiteratureBook ClubsReadingCommunityDigital Culture2026

About the Author

S

Taresh Sharan

PhD ยท IIT BHU

Research Scientist ยท Bangalore, India

PhD in Biomedical Engineering from IIT (BHU) Varanasi. Research Scientist based in Bangalore. Author of 200+ articles across AI, finance, photography, technical writing, careers, literature, and corporate ethics. Builder of the free Money and Health apps on this site.

Medical AITechnical WritingPhotographyPersonal FinanceLiterature
Full profile
Running an Online Book Club That Doesn't Quietly Die | Sharan Initiatives | Sharan Initiatives