In the space of only a few days, Michigan’s fall has gone from “crisp, sparkling, riot of color” to “gray, raw, endless rain,” meaning that we’ve now entered the season of meeting people for long afternoon talks in cozy coffee shops. That’s exactly what China media scholar Aynne Kokas and I did Wednesday afternoon, chatting for over an hour about Chinese tech, U.S.-China relations, academia, and writing. Aynne is working on her second book and I’m reading a never-shrinking pile of academic and trade nonfiction volumes, so she asked me a great question: If you could give nonfiction authors three pieces of advice, what would they be? In other words, what are the most frequent mistakes I see as a reader and book reviewer?

I hadn’t ever thought of this specific question, but I very easily came up with my top three tips and talked them over with Aynne. When I got home I shared those tips in a short Twitter thread, which has gotten a lot of engagement—enough that I decided to write up a slightly longer version of those tweets and post it here. For each tip, there’s a free option and one that requires some funds.

So here they are, my three top tips for nonfiction authors:

(1) Fact-checking is important. While some magazines, like the New Yorker, are famous for their rigorous fact-checking, book publishers have traditionally put the burden of this work on the author, with sometimes problematic results. As a writer, I fact-check to an admittedly compulsive degree, even things that I “think” I know. This is surely a manifestation of anxiety and imposter syndrome, but it’s also because as an editor and a reader I catch factual errors all the time. I get it: everyone is working too much, too quickly, and it’s easier to keep writing rather than stop and google something that you can retrieve from memory.

Except memory is imperfect, and some of those things in your memory? They weren’t right to begin with. As a book reviewer, I’m not going to come down too hard on someone who has mistyped and says that Mao Zedong died in 1975 instead of ’76 … but if an author writes that so-and-so was the first woman to do XYZ and I can very easily verify that’s not the case, I mentally mark down that author’s credibility a point or two. Enough of those errors and I reach two conclusions: (1) the author is sloppy, and (2) they don’t know this subject as deeply as I expect of someone who writes a book on it. The latter might not actually be true, but that’s what I think due to the former.

If I were reviewing that book, I would probably note the author’s carelessness and potential lack of expertise, and ultimately my review would get published and put on the internet for anyone to see. I’m generally a kind reviewer and not likely to get too snarky, but there are plenty of others out there who will write a nasty takedown based on factual errors, even small ones like in what year Mao died.

So don’t leave yourself exposed to those reviewers: fact-check. Even if you think it’s dumb, even if you’re 100% certain that every single word you’ve written is 100% correct—FACT-CHECK. If you’re an academic fortunate enough to have institutional funding, hire a professional fact checker; it will be money well spent.

(2) Two shorter chapters are preferable to a single long one. Long book? Fine. Long chapter? UGH. My personal reasons for saying this are both psychological and structural. Psychologically, seeing a 50-page chapter looming in a table of contents makes me dread it; before even reading a word of the book, I’m mentally preparing myself for a slog. Structurally, my time is very compartmentalized as I balance my writing work, day job, household tasks, and the rest of my life. If I’ve allocated an hour to reading before I head to the office and can’t finish one chapter in 60 minutes, I feel frustrated. I have a discrete block of time, I want to read a discrete block of material—a chapter—not get halfway through and then return to it later.

To make this less personal … I also tend to find that chapters longer than 20-30 pages suffer from a lack of narrative and analytical cohesion. They get bogged down in tangents or are stuffed with material that the author wants to include but doesn’t know where it should go. Academic writers might feel like they “need” chapters to be a certain length; as author Jonathan Chatwin commented to me on Twitter, “one of my tutors during my PhD said they thought short chapters felt ‘journalistic.’” I, obviously, have no problem with “journalistic,” but many academics would read such a description as a ding on the author’s scholarly bona fides. They need to get over that and embrace the short chapter.

As an author, you should be aware of a chapter’s length and pay attention to its structure and flow. What’s your main point? What evidence are you using to make it? Is your analysis concise and direct? Have you included an anecdote that you absolutely love but which doesn’t truly fit? Kill your darlings. Read your own work as critically and dispassionately as you can. Trade chapters with another writer and comment on each other’s drafts. If you have the resources to hire a developmental editor, they will help you work through these issues.

(3) Read the whole thing aloud. Whether I’m editing a draft for someone else or preparing my own writing for publication, my last step is always to read the entire work out loud. It’s not really fun. It can feel like an unnecessary extra task when I’m ready to be done with something and move on to the next piece. But the practice of vocalizing and listening to my writing always, always helps me identify awkward phrases, repetitious words, and “did I really write that?” moments. Taking the time to read your work aloud will make you a better writer, I promise. Make sure you have a large glass of water on hand, and break up longer projects into several sessions (this is also why it’s nice to have shorter chapters!).

A good number of Twitter colleagues endorsed this idea but noted that they prefer to use text-to-speech software rather than do their own reading. I also know of at least one academic author who hired someone to read their manuscript back to them—that might be a greater expense than most of us can handle, but depending on how you think and process language, it could be worthwhile.

So there they are—my top three tips for nonfiction writers. They might not be the magic key to publishing bestsellers, but all of them are important steps toward producing solid work that engages readers and stands up to reviewers’ scrutiny. You want them to focus on your argument and analysis, not get distracted by avoidable errors or clunky prose—or fall asleep in the middle of a chapter.

And did I read this entire post aloud before pressing “Publish”? Absolutely. Twice, in fact.