Questions to ask beta readers (by draft stage, with a copy-paste list)

Ask about their experience, not your craft

The cardinal rule: readers are unimpeachable on what they felt and unreliable on why. When a beta reader says 'chapter 8 dragged,' believe them completely. When they say 'chapter 8 dragged because you need more action,' take the diagnosis with salt — the fix might be cutting a subplot, not adding a fight. You own the diagnosis; they own the symptom.

This shapes how you phrase everything. 'Is the pacing good?' invites amateur editing. 'Where did you first check how many pages were left?' produces a fact you can act on. Ask for moments, reactions, and predictions — never for craft verdicts.

Early-draft questions: structure and story

On a first or second draft, you need to know whether the story works at all — so ask big-shape questions and explicitly tell readers to ignore typos and rough prose, or that's all you'll hear about. The goal is finding the load-bearing problems while they're still cheap to fix.

• Where did you first want to put the book down? Where did you actually stop reading for the day?
• Which character did you care about most? Least? Was there anyone you kept mixing up?
• What did you think was going to happen at the midpoint? At the end? (Predictions reveal what your setup is actually promising.)
• Was there anything you didn't buy — a decision, a coincidence, a reaction that felt off?
• If you had to cut one chapter, which one?

Late-draft questions: confusion, confidence, and the ending

Once the structure is settled, the questions narrow. Now you're hunting for local confusion, promises you forgot you made, and whether the ending lands with the force you built toward. Late-draft readers can also be your canary for anything that reads as unintentionally funny, confusing, or slow at the paragraph level.

• Was there any moment you had to reread to understand what happened? Where?
• Did the ending satisfy? Was it too fast, too slow, or missing a scene you expected?
• Any question the book raised and never answered?
• Did anything about a character contradict what you'd learned about them earlier?
• Would you read the next book? (Watch the hesitation more than the answer.)

Collect answers you can actually use

How you collect feedback matters as much as what you ask. The failure mode is the loose reply: an email that says 'the middle sagged a bit' with no page number, which you then spend an evening trying to map onto your manuscript. Feedback detached from the text is feedback you'll partially lose — every reattachment step drops information.

Two fixes. First, ask readers to note things as they read, not reconstruct afterward — in-the-moment reactions are the honest ones; memory smooths everything into 'it was good.' Second, use a format where notes arrive pre-attached to the passage they're about. This is how Scribegrove's beta sharing works: you send a private, watermarked reading link — no account needed on their end — and readers highlight and comment inline as they go, on their phone if that's where they read. Their notes flow back into your workspace anchored to the exact chapter and passage, and phone readers can even leave voice notes for the reactions that are easier said than typed. Nothing to reattach, nothing to lose.

However you collect it, close the loop the same way: triage every note into fix now, fix later, or disagree-and-ignore — and remember the old rule that if one reader says it, it's an opinion; if three say it, it's a problem.