Your Beta Readers Are Reporting Symptoms. You Need a Diagnosis.
Your beta readers came back. Three of them flagged chapter twelve, all of them used some version of the word "slow," and not one could tell you what to do about it.
So now what? You've got the thing every revision guide promises is the hard part (honest feedback from real readers), and you're sitting there with no idea which sentence to touch first.
Let's talk about the step those guides skip.
The gap nobody names
A beta reader is an instrument for measuring reader experience. That's what they're for, and a good one is worth more than a stack of craft books. When a reader tells you they got bored on page 240, they're giving you accurate, specific, genuinely valuable data.
What they are NOT giving you is a diagnosis. They can't. They were reading, not auditing. They felt the engine lose power and then reported the moment it happened, which is exactly their job... and the question of why the engine lost power belongs to you.
Nobody tells you that, though. The advice stops at "look for patterns," which leaves you holding a note that says "boring" and a chapter that looks fine.
Why "look for patterns" only gets you halfway
I want to be fair to the pattern-spotting advice, because it isn't wrong. Three readers flagging the same chapter IS meaningful, and one reader who didn't like your protagonist's name probably isn't. Sorting by frequency does real work: it tells you WHERE to look.
But knowing where to look and knowing what to fix are different problems, and only one of them gets covered.
The pattern gives you coordinates. Chapter twelve, three readers, all bored. Great. You still have to open chapter twelve and determine what's structurally wrong with it, and "bored" is not a structural category. There's no revision you can make to a scene called "make it less boring."
That's the whole gap. And the way across it is to stop treating the reader's word as a description of the problem and start treating it as a pointer to the problem.
Three symptoms and where they usually come from
Reader responses cluster. Not perfectly, and not exhaustively (this is a working map, not a taxonomy), but enough that you can usually narrow the search before you start reading.
Boredom and disengagement. This is the big one, and it's almost always structural. A scene where nobody wants anything, a scene where whatever they wanted isn't being blocked, or a scene that ends in the same place it started. Readers don't experience any of that as "no goal present." They experience it as their attention wandering to their phone.
Confusion. This usually points somewhere else: cause and effect that doesn't connect across scenes, context the reader needed two pages earlier, or narration that keeps changing how close it stands to the character. Worth saying plainly that I don't have a single craft authority to hang the cause-and-effect half of this on. It circulates widely and I believe it, but I'm not going to dress up common sense as doctrine.
Disbelief. The "I didn't buy it" note tends to come from a decision the character makes without the groundwork to support it, or a consequence out of proportion to what caused it. Same caveat as above on sourcing.
Two of these have real craft frameworks behind them, and those two are where this gets useful.
Audit one: does the scene actually work?
Dwight Swain laid out scene structure in Techniques of the Selling Writer back in 1965, and it has survived sixty years because it's a test rather than a theory. A scene, in Swain's terms, runs on three parts: the character wants something specific and immediate, something opposes it, and the scene ends in what he calls a disaster. Not a catastrophe necessarily. A setback. The situation at the end is not the situation at the start.
(You'll see "scene turn" used for roughly this idea all over the place, including by me. Fair warning that it's broader craft usage and not Swain's word. His word is "disaster," and if you go reading him it'll save you some confusion.)
He pairs it with the sequel: reaction, dilemma, decision. The character absorbs what just happened, weighs bad options, picks one. Skip the sequel and readers feel yanked along without knowing why they're unmoored.
So take your flagged passage and ask three questions, yes or no, no partial credit:
- Does the viewpoint character enter this scene wanting something specific? Not "wanting to be happy." Wanting to get the keys, win the argument, avoid the conversation.
- Does something get in the way, inside this scene?
- Is the situation at the end different from the situation at the beginning?
Three noes and you've found your problem. One no is often enough. The useful part is that every answer points at a specific repair, which "boring" never did.
A detailed walkthrough of Swain's framework lives on September C. Fawkes's blog, and the Wikipedia article on scene and sequel cites his page numbers if you want to go to the source.
Audit two: how close are we standing?
The second framework is John Gardner's, from The Art of Fiction (1983), and it covers ground Swain doesn't touch.
Gardner describes psychic distance as the felt closeness between reader and character, and he maps it as a continuum of five levels. At the far end the narration is wide and reportorial, the camera up on a crane, telling you about a man in a town. At the near end you're inside the character's thoughts as they occur, with no narrator standing between you and them. In between are three gradations of increasing intimacy.
Gardner's point, the one that matters for your chapter twelve, is that moving between those levels is ordinary craft when you mean it and a problem when you don't. An accidental jump from level two to level five and back again produces a very particular reader experience: something feels off, the prose seems competent, and nobody can say what's wrong.
Sound like a beta note you've received?
The audit is mechanical. Read the flagged passage and mark each paragraph for roughly where it sits on that continuum. Don't agonise over whether something is a two or a three. You're looking for jumps, places where the narration leaps several levels in consecutive sentences with no reason to. The University of Nevada Reno writing centre has a clear explainer that walks all five, citing Gardner.
One caution. Gardner is talking about CARELESS shifts. Deliberate modulation is one of the best tools you have, and this audit is not an instruction to flatten your narration into one register.
The whole protocol, start to finish
Put together, here's the pass:
- Go through every note and mark the passages that got an emotional response. Bored, confused, lost, didn't care, didn't buy it. Call these your symptom sites. Fix nothing yet.
- For each site, find the scene boundaries. Where does this scene start and stop?
- Run the three Swain questions. Goal? Opposition? Changed situation? Write down which ones failed.
- Run the distance check. Mark the paragraphs, look for unintended jumps.
- Build a two-column list. Symptom site on the left, structural findings on the right. One site can produce several findings, and often does.
- Sort by overlap. A passage with three findings and two readers flagging it is where you start. A passage with one minor finding and one reader is where you stop, or don't bother.
- Revise against the finding and not against the wording. This is the part people skip. The reader said "boring." Your finding says "no goal, and the situation doesn't change." You're now installing a goal and a consequence. The word "boring" has done its job and you can let it go.
That last step is the reframe this whole piece is built on. Once you've got a structural finding, the reader's language stops being useful and starts being noise. You are no longer responding to notes. You're making craft decisions, which is a much better place to be working from.
What this doesn't cover
This is a scene-level and passage-level tool.
It won't fix prose. Sentence rhythm, word choice, dialogue that clunks... different pass, different skills. It won't diagnose a premise that isn't working, or a character arc that fails across the whole book, because those problems don't live inside one scene where this audit can see them. Act structure and whole-manuscript pacing are their own job.
And some feedback genuinely is preference. A reader who wanted more romance in your heist novel has told you something about that reader. Running a structural audit on it will turn up nothing, which is itself a useful answer and a decent way to tell the two kinds of note apart.
Where FirstReader fits
The protocol above is designed to need nothing but your manuscript and a notepad, and I'd rather you ran it by hand first, because doing it yourself a dozen times is how the questions become automatic.
Where it gets impractical is scale. Running two audits across ninety scenes by hand is a long week, and the honest truth is that most people quit around scene fifteen.
FirstReader analyses a manuscript at scene, chapter and book level against craft doctrine, and it scores Scene Structure as one of its six craft dimensions, defined as goal-conflict-disaster architecture, scene turns and narrative purpose. Narrative distance sits under the Point of View dimension, where the findings it reports include psychic distance breaks. So it can run the same two passes over a whole draft and tell you which scenes have no clear goal, which ones end where they began, and where the distance to the character moves around.
What it does not do, and I want to be precise here, is read your beta feedback. It has never seen your readers' notes and it can't tell you why somebody was bored. It analyses the manuscript, reports what's structurally there, and the translation from "three readers flagged this" to "here's the cause" stays your act of judgement. Use it as a second pass on scenes you've revised, or as a way to find the symptom sites your readers didn't happen to mention.
If that sounds useful, you can try it on a manuscript.
What you actually do next
Pull up the feedback. Mark the passages, don't read the adjectives too hard, and ask the questions.
You'll probably find that two thirds of your notes collapse into four or five structural problems that recur, which is a far smaller job than the pile of comments suggested when you opened it. That's the usual outcome. Readers flag symptoms in a dozen places, and the causes turn out to be a handful of habits.
Final words: the note that stings worst is usually the most useful one, and "I was bored" is not an attack on your prose. It's a measurement. Go find out what it measured.
If you enjoyed this, please leave a comment below to let me know. If you DIDN'T enjoy it, well, I'd like to hear from you too!
If you haven't already, please feel free to subscribe so you don't miss anything.