Well, I mean I don't want to judge all startups.

In fact if you take any of this offhand advice as gospel, well I wouldn't.

Maybe I should have scattered "probably" and "often" about the place.

1: ------------------------------------------------------------

When you read the book, ole Fred is not a fan of large engineering units. Forget two pizzas (Bezos is from New York, two pizzas for him is ~10 people).

Fred's view is actually:

1x captain/tech lead,

1x co-pilot,

1x ops guy equivalent,

1x business analyst equivalent who can be sent to meetings (he worked for IBM).

(And then two secretaries. And an copy editor for the documentation, because no internet.

No internet! The documentation must be written!

The lead developer needs to follow my favorite advice of all time:

Can't write? Imitate Hemingway.

And a full-time librarian for the punch cards. And a full-time employee just to maintain the minicomputer that was your debugger.

Because hey, the book was written in 1975.

The internet doesn't exist, email doesn't exist, the engineers almost certainly can't touch type. And anyone without an attractive female secretary was nobody.

The pimply faced youth's job as config wunderkid and unofficial court jester fills most of those roles these days)

2: ------------------------------------------------------------

If you think of all the amazing software that has been made by one person...

How many people do you need to make amazing software?

What are you even making?

Depends on the product what you actually need, but I've personally watched more than one successfull software product launched with a more or less this resourcing. Not exactly this recipe but very much in the ballpark.

The assistants aren't assistants - you hire them with minimum pressure, but try still to find the most talented junior you can find.

There is no joke, that's correct.

You won't need more than 2 engineers at the beginning for 99% of the startups. Almost nobody is making anything technologically significant - and if they do they raise millions and act like a disorganised mid company.

In order to do your typical backend + frontend work you won't need much more, unless your developers can't code.

It's heartening to hear how remote work is helping to even out global wealth inequality.

Oh, I could be wrong on the number. It's extremely low though, certainly in the teens.

As a person they're very opinionated about not working for a "product company." And have worked at the same company their entire tech career after bailing out of some traditional engineering discipline (I think 6 yoe?).

Sweetest person ever. Never really followed up that conversation when I asked them if they were considering remote.

Brazil is the beachhead in many ways. Huge market, easier to sort out all the employment stuff because it's a path that's much more well trodden.

If you still want to get an edge these days, you have to be prepared to go to the global platforms that give you access to devs in the really obscure places.

Nobody is hiring from Trinidad and Tobago because of the legal barrier to entry. But you can if you are clever with Upwork and similar.

I imagine the answer to that is, I've simply never heard of the US government buying from a software vendor with less than 500 employees.

Surely it never happens?

Total Cost of Employment: https://support.remote.com/hc/en-us/articles/4406888078733-W...

"delta" = difference.