Testing the Good Writing app

An editing / feedback tool designed around Good Writing: 36 Ways to Improve Your Sentences

After Anne Lamott posted about the goodwriting.app on Substack, I decided to take a quick look to see what an editing tool built around the book Good Writing: 36 Ways to Improve Your Sentences by Neal Allen and Anne Lamott would look like.

For the book, Neal decided to write down his thoughts on some of the writing rules that he had learned from other writers and applied to his own professional work. Anne decided to throw in her two cents, mostly making fun of Neal, while adding her unique perspective.

It’s a very good read.

The way Anne described the Good Writing app in her post made me think it might be a useful tool. It could look at some of my writing, point out potentially weak parts and explain why it thought they were weak. It doesn’t write for you or suggest what you should write (specifically) which should reduce objections to its use for some of those opposed to writing with AI.

Because I own the book and I am a fan of both Anne and Neal, I would have checked it out regardless. But, given all the hullabaloo around Substack’s addition of Pangram AI detection, I found it even more interesting.

Regarding AI in the big picture, and its use in writing, I won’t go deep into my personal opinion because it would take way to long too explain my position, and it would just come off as yet another (mostly pro) AI rant. When it comes to consuming the writing of others, anything goes. If I enjoy something, I enjoy it. When it comes to writing, my feeling is that if you aren’t experimenting and incorporating AI into your writing process you’re doing it wrong. I do love all your 100% human content. Things that you scribbled down in your chapbook using a quill. If that’s your process and it works for you, great. If you use AI as part of your process, I am happy about that too.

I use AI for a lot of reasons, many of them work related, but I also have long discussions with any number of LLMs about writing in general, and specifically, my characters, how I feel about them, and what I want to express through them. The best thing about AI for me is that it’s instant on. There is always someone there to “discuss” things with me, and occasionally ask a question I hadn’t thought to ask.

So, I’m fine with AI assistance for work, self-therapy, and writing.

As a trial run of the tool, I took something I had written fairly recently, was important to me and ran it through the app to see what kind of feedback it would provide.

This is the opening of a chapter from a novel I have been working on for the last month or so. The text has been color coded to show things the tool flagged. It’s reasoning is shown and discussed below.

---

It was an unusually hot day. What had been a mild, rainy June began to simmer as the days slid into July. Kyoko and I sat by the pool, our loungers tucked into the shade of an umbrella to escape the midday sun.

Kyoko sat next to me, hands resting on her lap, a small bead of sweat forming at her left temple.

She wore an oversized yellow button-down over her swimsuit, the sleeves rolled twice at the wrists. The shirt no longer belonged to anyone in particular—sometimes it was mine, sometimes hers. She must have been unbearably warm beneath it, but she seemed more willing to sweat than take it off.

“It’s too hot for that,” I said.

“I know.”

She made no move to unbutton it.

---

The writing reflects the style, pace and language I am going for fairly well. It took me a long time to find this voice. The typical kind of writing process where you pull out your hair and most days you want to quit. Even though this is still a draft, I feel like it is close to what I want.

So, what did the goodwriting.app suggest to make it better?

It was an unusually hot day.

Rule 2: Question “Being” and “Having.

This pins a label on the day instead of letting the reader feel the heat. What does that heat actually do—to the air, to the skin, to how people move and breathe? Let the scene carry the temperature.

Great. I get it. But I feel really comfortable with this sentence and the fact that it opens the chapter.

Kyoko sat next to me, hands resting on her lap,

Rule 9: Question Transitions

You've already placed Kyoko in the scene with you by the pool. Saying she 'sat next to me' announces a spatial relationship your reader has already absorbed from the opening image. The transition makes the description feel like a checklist rather than a continuous moment.

Here, the tool makes a pretty good point. Although not explicitly stated, the reader probably assumes they are sitting together. What’s great is that the feedback shows that the tool is aware of the context and positioning of the characters.

I could easily see me changing this line to:

Kyoko rested her hands on her lap.

However, I don’t hate the narrator being explicit by saying ‘next to me’.

The shirt no longer belonged to anyone in particular—sometimes it was mine, sometimes hers.

Rule 6. Prefer Anglo-Saxon words

This formal word (particular) distances the reader from the casual intimacy of the moment. The shirt's shared ownership is intimate and everyday; this word makes it sound like a logical statement rather than something lived and felt.

I took the long way around to make this statement. I like it. Here are some alternatives.

The shirt belonged to whomever was wearing it.
The shirt was mine, or hers, depending on the day.

I’ll have to think more about this one.

The shirt no longer belonged to anyone in particular—sometimes it was mine, sometimes hers.

Rule 13. Dress up “This”

The reader has to reach back and decide what 'it' refers to—the shirt itself, or the state of belonging to no one? Naming the referent would lock the sentence into focus.

This is a good rule and I constantly catch myself using ‘it’, or ‘that’, or ‘this’ instead of making it clear for the reader. Here, I don’t think ‘it’ should be flagged. If it’s not 100% clear that ‘it’ refers to the shirt then the reader has stumbled somewhere along the way (and of course, that’s my fault.)

The shirt no longer belonged to anyone in particular—sometimes the shirt was mine, sometimes hers.

Is this better? I don’t think so, but you might disagree. Perhaps this is a case where the entire sentence is the problem and an alternative, simpler construction solves all problems.

She must have been unbearably warm beneath it,

Rule 6. Prefer Anglo-Saxon Words

This Latinate word creates a layer of abstraction between the reader and the physical sensation you're describing. The moment calls for something direct and visceral—the reader should feel the heat pressing down on Kyoko, not observe it from a distance.

Rule 6 again. I like the feedback, however, this is the narrators observation, so inherently, at a distance. Anyway, Kyoko doesn’t seem to mind, otherwise she would take off the shirt, and we wouldn’t need any of this.

She must have been unbearably warm beneath it,

Rule 13. Dress up “This”

The first 'it' is clear (the shirt), but the second 'it' in 'take it off' asks the reader to hold two possible referents in mind—the shirt or the heat. Naming what she refuses to remove would sharpen the image.

Okay, Rule 13 again. However, I like the feedback (generally speaking.) Again, I disagree. it = shirt should be fairly obvious. Still, I would consider changing the second occurrence of ‘it’.

She made no move to unbutton it.

Rule 1. Use Strong Verbs

"Made" is a generic causative verb that asks the reader to do extra work. What does the absence of action actually *look* like here? Let a more specific verb show us the stillness or resignation directly, so we feel it rather than having to infer it.

Rule 1, Use Strong Verbs is my favorite rule, and it is very practically applied and is an easy way to strengthen your writing. That’s probably why it is rule #1.

I like this sentence. Much like the opener, “It was an unusually hot day.” I see no reason to change it. If I did, similar to my thoughts around the ownership of the shirt, I would reconstruct the entire sentence, which would make the point about using a Strong Verb inapplicable.

Perhaps:

She made no move to unbutton the shirt.

Probably better.

She seemed to be in no hurry to unbutton the shirt.

That’s a fine sentence. Longer though. Perhaps ‘seemed’ is not a strong enough verb though?

---

My conclusion is that the tool is very easy to use. The feedback has some depth to it. It doesn’t feel like a simple expert system flagging keywords, and every suggestion is tied back to one of Neal’s rules.

For me, though, that’s not the interesting part. The interesting part is that it got me thinking about what I’d written, why I’d written it that way, and whether I actually wanted to change it. Sometimes the answer was yes. Sometimes it was no. That’s exactly how I think an editing tool should work.

One thing I noticed is that the tool is often better at identifying a sentence worth reconsidering than nailing the best revision. That’s probably a feature rather than a bug. For me, once I decide a sentence isn’t working, I rarely tweak it; I just rewrite it from scratch.

In the end, I may keep every sentence in this sample exactly as it is. I may rewrite half of them when I get to the revision stage. At the moment, the ownership of the shirt probably feels like the weakest area, but even there I’m not convinced I should change it.

Ultimately what I like (so far) about goodwriting.app is that tt didn’t try to write for me. It simply gave me another reason to stop, think, and make a conscious decision.

For a writer, that’s a pretty useful thing to have.