Beta readers: how to choose them and what to ask
Beta reading is the moment a manuscript stops being only yours and meets, for the first time, eyes that don't already know how it ends. It's also one of the most misunderstood steps in revision: sent to the wrong people, with the wrong questions, it produces noise instead of useful information.
What a beta reader does, and doesn't do
A beta reader reads the manuscript the way any reader would, and reports their experience: where they got bored, where they didn't believe a character's choice, where they lost track of who was who. They aren't an editor: they don't need to know how to fix a problem, only how to recognise and describe it.
Confusing the two roles is behind most of the disappointment. A beta reader who proposes editing solutions is doing different work — useful, but different — and should be treated as such, not as the final verdict on what to change.
Who to pick: readers of the genre, not just people you trust
The most important criterion isn't "who cares about me" but "who genuinely reads the genre I've written." A strong mystery reader senses whether the clues work; someone who never touches a mystery has no yardstick to tell you, no matter how well-meaning they are.
Friends and family can be excellent beta readers if they read the genre and are willing to be honest, but the emotional bond is often an obstacle: some soften their judgement to avoid hurting you, others go the opposite way and are harsher than necessary precisely because closeness feels like permission. A mixed group — a few people close to you, a few genre readers who don't know you — balances both risks.
Three or four, not ten
More beta readers doesn't mean more clarity. With three or four opinions it's easy to see where they converge — that's the real problem — and where only one flags something the others missed, which needs weighing but not necessarily following. With ten opinions, the signal drowns in noise, and every page risks a contradictory critique from someone.
If a detail is flagged by three readers out of four, it's a problem. If it's flagged by one out of ten, it might just be their personal taste.
Specific questions beat generic ones
"Did you like it?" almost always produces a polite, useless answer. Useful questions are anchored to a precise point in the text and a concrete reading experience: "at what point did you think about stopping?", "where did you start not believing [character]'s actions?", "what did you expect to happen after chapter ten, and what actually happened?"
Pacing questions work best asked by section: "which chapters did you read faster, which slower, and why." Character questions work best asked about a specific moment: "in this scene, did their behaviour feel consistent with what you knew about them?"
When to send the manuscript: never mid-draft
Sending chapters as you write them looks like a way to get quick feedback, but it almost always damages both the book and the relationship with the beta reader: a judgement on a third of the book can't assess how the setups pay off, and it risks pushing you to change direction before the story has had a chance to show where it wanted to go.
The right moment is after a complete draft, ideally after your own first revision pass — sending the very first rough draft wastes readers' attention on problems you'd have caught yourself anyway.
Handling conflicting opinions
When two beta readers give opposite feedback on the same point, the temptation is to look for a middle ground. That's often not the right answer: it's worth asking whether the two are flagging the same symptom from different angles, or whether one of them simply has a different genre taste than your ideal reader.
The final call stays yours: beta readers give you data, not decisions. A novel written to please every beta reader stops having a voice of its own.
Keeping feedback from several readers organised
With three or four beta readers commenting on different chapters, at different times, with observations that only partly overlap, keeping track of who said what quickly gets complicated. In NovLore you can share an expiring read link with your beta readers without giving them access to the editor, and note their comments in the dashboard next to the scenes they refer to, so feedback stays tied to the exact point in the text instead of scattering across emails and messages.
Frequently asked questions
How many beta readers do I actually need?
Three or four are usually enough to spot recurring problems without drowning in contradictory opinions. Fewer than three risks mistaking personal taste for a structural problem; more than five or six becomes hard to manage without losing the signal.
Should I pay beta readers?
It isn't the norm, but some writers offer a symbolic fee or a swap — reading the beta reader's own manuscript in return. What matters most is being clear from the start about timelines and expectations, paid or not.
What do I do if a beta reader doesn't finish the book?
That's valuable information in itself: ask where they stopped and why. Abandoning a book partway through, especially if it happens to more than one reader at the same point, almost always points to a pacing or interest problem in that section, even if no one puts it into words as criticism.