Thread by @HeyChelseaTroy

Tech books exhibit a strange cost bell curve relative to quality.

Expense-it-to-prodev priced books are consistently fair-to-middlin'. Accessibly priced books, a standard deviation above or below that. Free books, either TRULY shite, or the best tech writing I've ever read.

My hypotheses on why come from my experiences:

- planning books with big publishers
- hearing from published author colleagues
- getting pitched on self-publishing
- self-publishing for reasons totally unlike the pitches

Here they are, in all their half-baked glory:

1. Books from big publishers

I won't name names, but if you've been around tech, you know who this is. These are the places with the highest price point. It's that high because they expect people to expense it to their employers. These places have a lot of name recognition, and

...they base their decisions about what to publish on industry buzz and tenured authors' recommendations.

These places have a price point high enough that they can pay decent editors. Authors provide content and name recognition, but aren't expected to write well from jump.

That puts a lower bound on the quality of these books.

But also, these places have a format. They want 250-300 pages. They want three sections. Please include chapters on the following buzzwords. The editors help shape the author's words to "The Company Voice."

This...

...results in some antipatterns that you can see in these books, if you know to look for them:

Example: a "kitchen sink" chapter that is terrible, disorganized, uninformative, and a massive departure from the rest of the book. But it wins buzzword bingo, as requested.

That kind of thing puts an upper bound on the quality of these books.

When we step down a level to "accessibly priced" books, we're usually talking about scrappier publishers without the name recognition or chutzpah to charge something an individual wouldn't pay on their own.

These places decide what to publish by looking up the sales numbers of the expensive-tier publishers, grabbing the top-selling ten titles, and commanding authors to write direct competitors that address the feedback on the originals from Amazon reviews.

The editing at these places is spottier. MUCH spottier. It's hands-off and not communicative. The editors aren't necessarily fluent in the spoken language that the book is written in, and they almost CERTAINLY aren't fluent in the programming language the book is about.

This shoddier editing also means less enforcement of brand voice or book structure.

That allows authors who are already good writers to bend their book in a higher-quality direction, but it also fails to correct for weak writing. Hence the standard deviation above and below.

Finally, we have free (which are usually self-published or company-published) books. Here's why I think these are either the best or the worst.

The most commonly HAWKED reason to self-publish for free is "lead generation". I.e., create collateral for someone's contact info...

...such that you can try to get them to buy your paid stuff later.

Frankly, a lot of lead generation type stuff reads like something the author shat out in haste.

Or it withholds the good info because the author wants to charge for their 'secrets'.

Or it's basically an ad.

Then there's the other kind of self-published stuff.

Lemme acknowledge up front, the common theme among REALLY good free or self-published writing is privilege.

This stuff is written by people who get paid enough at their day job that they don't need their writing to earn.

As much as I hate to say it, I think you could probably predict the quality of free tech writing with a term where the numerator is the number/prestige of companies the author has worked at, and the denominator is the number of times they mention those companies in the book.

This isn't because FAANG techies are better writers; it's because FAANG techies get paid enough to not care if their book makes money

(the denominator term is because IME books where the author can't stop mentioning that they worked at Google also trend, uh, not that good)

ANYWAY, so they're not doing it for money.

FWIW, I am also in this boat. I don't really have lead generation pieces. I have an email list to which I send cleaned-up rollups of blog series I already made available to the world for free. If I never get paid to write, that's fine.

This is a brief aside but to clear up confusion: the reasons I write:

1. Cement my knowledge of the topic
2. Establish the robust external validation of readers so that my self-image as a programmer doesn't depend on any one client/manager
3. Help other people learn the topic

People who write for reasons like that, rather than needing the writing to earn, tend to be pretty fastidious about writing well.

Also, thanks to the privilege part, those people can pay skilled editors. They can hire illustrators. They can pull out all the stops.

(and btw...

...they don't even necessarily need to, because they also have strong networks of skilled people who will help for free).

Also btw, in the end they do make money indirectly off the writing because someone will pay them $600 to give a podcast interview or $5k to do a talk. But this income is both delayed and not guaranteed, and it's not, IME, why these people write.

IME they're writing, it turns out, for some personal reason.

People who do this enjoy writing, which correlates with more practice, which causes better writing.

Plus, personal reasons typically set the bar higher than "get something out the door that somebody will buy."

So I want to be careful in this thread to clarify that I'm not equating better writing, or free-er writing, with more worthy or moral writing. As we've elucidated, there are structural inequalities associated with who "gets" to write what, and with what support.

But...

...for folks who possess limited learning budgets and need to spend it wisely—and by the way, that's in terms of money AND in terms of time and energy...

it's helpful to consider the non-linear relationship between cost and quality and stronger correlates for effective writing.