Beta reading is the cheapest quality step available to a self-published author and the one most often done badly, usually by sending the wrong draft to the wrong people and asking the wrong question.
When to send it
After the second draft, not the first.
A first draft invites feedback about things you already know are broken, which wastes the one thing a beta reader can never give you twice: a first reading. Fix the structure yourself first (self-editing covers that pass), then hand over something you believe in.
Signs it is ready: the structure is settled, nothing is a placeholder, and you would be comfortable if a stranger read it without explanation.
Who
Five to eight is the working number. Fewer than three and one opinion dominates; more than ten and you get contradictions with no way to resolve them.
| Reader | Good for | Watch out for |
|---|---|---|
| Your actual target reader | The only feedback that counts | Hardest to find |
| Someone who reads the genre | Convention, pacing | May want a different book |
| A subject expert | Errors of fact | Wants more depth than readers do |
| A friend who reads a lot | Honesty, if they can | Kindness, if they cannot |
| Another writer | Craft | Rewrites it as theirs |
Avoid your family, unless they have a demonstrated ability to be unkind to you. Not because they lie, but because you will not believe them either way.
Where to find them: readers of the newsletter you have been building while writing, genre communities and forums, local writing groups, and, for non-fiction, people who have the problem the book solves. That last group is both the best and the easiest to reach, because you probably know several.
What to ask
"What did you think?" produces "I liked it", which is worth nothing.
Ask about their experience, not their judgement:
- Where did you stop reading? The single most valuable question. Not *whether* they stopped, but where.
- Where were you confused? Page or chapter.
- What did you skip? People skim; skimmed sections are cuttable sections.
- Was there a point you nearly gave up? Usually the middle.
- What did you expect that never came?
- Who would you describe this to? Tells you whether the positioning works.
- What do you remember, a week later? Ask this a week later. It is the real test.
For non-fiction, add: could you do the thing afterwards? That is the whole promise of the book.
How to send it
- A clean EPUB or PDF, not a Word file with your comments in it. Making it feel like a real book gets real-reader behaviour.
- A deadline. Three to four weeks. Without one, half never finish.
- The questions attached, not in a separate email.
- No explanation of the book. If it needs one, that is the finding.
- Say thank you in the acknowledgements, and mean it. Front and back matter is where that goes.
Sending a proper file matters more than it sounds. A reader who has to scroll a manuscript reads like an editor; a reader with something that behaves like a book reads like a reader, which is what you are trying to observe. bukpress produces both formats from a Markdown draft, so a beta copy is a build rather than a task.
Reading the feedback
The rule: one reader is an opinion, three readers is a problem.
- Everyone stumbles in the same place → fix it, whatever they say the cause is.
- One person hates something everyone else liked → they are not your reader.
- Contradictory notes on the same passage → the passage is unclear rather than wrong.
- Nobody mentions the thing you were worried about → you were worried about the wrong thing.
Wait a week before acting. The first reaction to criticism is not a good editor, and roughly a third of what stings on Monday is obviously right by Friday.
What beta readers are not
They are not proofreaders, so do not ask them to catch typos, because it changes how they read and they are not good at it. They are not an editor. And they are not reviewers: a beta reader who loved the book is not obliged to post about it, and asking converts a favour into a transaction. Getting book reviews is a separate effort with separate people.
Frequently asked questions
How many beta readers do I need? Five to eight who finish. Recruit ten to twelve, because some will not.
Should I pay beta readers? Usually not: unpaid readers behave more like real readers. Paid beta services exist and are closer to a light editorial report, which is a different and also useful thing.
What if a beta reader is harsh? Separate the tone from the content. Harsh delivery of a real problem is still a real problem, and the least pleasant reader is often the most useful one.
Should I send the whole book or chapters? The whole book. Chapter-by-chapter feedback cannot tell you anything about pacing, structure or whether people finish, which is most of what you are asking.
Do I have to act on all the feedback? No, and you should not. Act on what several readers hit independently; note the rest and move on, or you will edit the book into a committee's version of itself.