Item — HN Reader

The thing that I think makes this letter great, and the reason I upvoted it on a tech site, that I haven't seen anybody else mention yet is this: it's a universal point.

Movies aside, I think his point applies to a lot of other industries, "If I worked at a donut stand, and I kept fucking up donuts, I'd be fired. Even if I made a tiny decent one every now and then, it doesn't matter. I'm gonna get fired." I have found this myself in tech, that often those held as icons of genius are often held to lower standards and have the potential to do much more damage than those at the bottom.

For example, a startup may have an awful product yet the pressure always trickles down to engineers to make it 10% faster or get in trouble. Yet so often it seems not to be that 10% delivery rate that separates a unicorn from a failure. It's a truly novel idea, a great design, and architects who can implement that vision in a stable and smooth way.

1 comments

My experience made me identify with this letter very closely-- in broad generalized terms, you often have "management" which is a bunch of "business" people who are directing "creatives" (whether software engineers or movie artists) based on the management's perception of the state of the market.

Unfortunately, there is always a tendency to believe your own PR, believe you know more about things than you do, and to believe that, because you're on top and they are on bottom, you know better.

You see this in startups also when the "management" is a 20 year old CEO who just graduated business school, as much as when he is a 60 year old who has worked in the industry (but whose background is not in actually making things).

This is the big gulf between the makers and the managers.

And I think the correct response is, as makers, to make sure we are managed by other makers.

An engineer in the CEO role can manage a VP of marketing, or a VP of sales or a VP of engineering fine. Find a VP of marketing who wants to do customer discovery, find a VP of sales who can close (and incentivize him to do so, if he's not hungry don't hire him) and you'll do better, in my experience, than a CEO who doesn't understand how software is made (and subsequently keeps inadvertently undermining his own products as a result.)

Apologies if this seems like a bit of a rant- this is a comment on a letter that is a rant, though-- but I've seen this over and over and over in the past 30 years working for technology startups.

I would say having a technical CEO has a positive order of magnitude improvement on the outcome, and liklihood of a liquidity event and the size of that liquidity event, in my experience.

It lowers risks many ways, including increasing retention of your most valuable employees. (One of the key things that business people can't seem to get over is the idea that commoditizing employees leads to hiring what Paul Graham calls "blub" programmers, but that if you hire great programmers you can't treat them like cogs or they will leave.)

The problem is that (general intelligence aside) management skills are orthogonal to engineering/art/other creative work. This means that it takes a lot of work to find and train creators to be managers. On top of that (at least in engineering), a large fraction of the quality engineers I know actively do not want to be managers.

If we haven't seen as many screw-ups due to lack of managerial skills among engineer-managers as we've seen screw-ups due to lack of engineering savvy among MBA-managers, it's at least partly due to the fact that there are a lot more of the latter than the former.

Not in my experience. Or put another way, if you take someone with an MBA or a masters in engineering, neither of them intrinsically has more management skills or training. (IF colleges are teaching "management skills" they are not doing a good job of it.)

Yes, many engineers do not want to be managers, and unfortunately, too many non-engineers want the power. This is a good thing about engineers and a bad thing about those non-engineers.

I've seen lots of good engineers who manage, but almost no business people who are good managers of engineering.

I know that what you're saying is a popular perspective, but so is the idea that we should let wall street manage our retirements and real estate is always a good investment. Some of this perspective is self serving.

"I would say having a technical CEO has a positive order of magnitude improvement on the outcome, and liklihood of a liquidity event and the size of that liquidity event, in my experience."

Interesting assertion. Lucky for us, there are plenty of statistics available. So instead of going with gut feeling, do you have any sources to back this up? (Genuine question - I'd love to find out whether this is actually true).

"Unfortunately, there is always a tendency to believe your own PR, believe you know more about things than you do[...]".

While you make this point about CEOs, how do you know you're not talking about yourself as a maker?

Note: I'm not against your point, but I think the hubris of "makers" vs "everyone else" is just another example of people thinking their ingroup is the "right" and "successful" one.

I'm not aware of any statistics that qualify the CEOs of startups based on their background. Maybe you could do a rough pass based on degree program. My assertion is purely based on my experience, but it's not just anecdote- I've worked for 2 dozen startups and have visibility in to 3-4 times that.

> how do you know you're not talking about yourself as a maker?

This is a good point, and it gets to what my real point is: Culture. The issue is not lack of skills or integrity it's a different culture.

The culture of making things is different from the culture of running a 1950s era light manufactory. The latter is the "management" that most business people are taught. IT's woefully out of date. And it's predicated on the idea that you're stamping out the same parts over and over again. Tech is cutting edge and thus more of a craft than a manufacturing enterprise.

Engineers know this and so they know what to optimize for. Business people it seems can't or won't accept this.

They always want to know the date. IT's like its the only thing they care about. IF you ask an engineer to estimate something where the spec is a vague two sentence description they won't over-estimate themselves and think they can predict the time. Business guys, for instance, never seem to understand why you won't give them an estimate (which becomes a deadline, they don't seem to understand that estimates are %50 wrong if the are given correctly.)

this is because their culture is coming from a land where it takes exactly the same amount of time to stamp out a widget and predicting how long it will take to make 50 widgets is a no-brianer.