Your definition of a 10x engineer is a pile of crap.

Notes from The Revelry Blog


I’m going to assume that almost anyone reading this is already fairly familiar with the term. Somewhere along the line a couple years ago, the whole industry started buzzing about “10xers” and how you should hire them and only them. And then went on and on about what all the supposed attributes are of an engineer who contributes 10x the average.

And a pile of follow-up content about how to be a 10xer. Be creative, be independent, be curious, be energetic, be leaderlike. Be some combination of superb attributes (at least half of which would be totally attainable if you could pick your own personality traits at birth) so that you can do 10x the work of the nearest plebieian.

And while I don’t doubt at all that natural talent is extremely significant, nor would I suggest that work ethic alone can fully compensate for a lack of it, there’s just something deeply irritating about the whole conversation. Namely, it doesn’t really give the supposed rest of the class any sort of near-term goal. It merely says, “This is how far you are from being worth my time.”

I think the heart of the problem is that we’ve got x all wrong.

Typical definitions of x

If 1x is the output/quality/etc. of one average developer, then 10x is obviously ten times that. Makes sense. Sound elementary school math. No debate to be had.

On the other hand, many would argue that those people don’t exist. You can’t excel at everything. Sam Gerstenz attempts to sort of correct the concept in this Andreessen Horowitz blog post:

A “10x engineer” is, afterall, an incredible misnomer. A 10x engineer is not necessarily 10 times more productive — they are just “next level” better engineers who in some contexts are 1.5x and sometimes 100x better, depending on the difficulty of the task and leverage of the outcome.

At least this is a bit more grounded. But as Robert, one of our senior engineers, sarcastically said to me, it’s like the Valley a couple years ago just discovered “this totally new and novel concept of ‘experience’ and ‘specialization’.”

Sam kind of brought it back to almost-earth but also kind of betrayed the term’s worthlessness: If it’s all context-based, the implication is that you should ideally only hire precisely the perfect employee for the particular context. Which isn’t worthless because it seems obvious. It’s worthless because in a world of infinite contexts, there’s no more guidance to give. There’s no path toward 10x. Not toward being it. Not toward finding it.

Literally the only advice I can take away from this is that at the end of an interview process I should ask myself, “But is this person a 10xer?” Barf. Okay, let’s get on to what I think is the fundamental problem that plagues the whole concept:

In every single definition of a 10xer I’ve seen thus far, not only does the developer possess superior attributes, but x belongs to that developer and that developer alone. A developer whose own ability and output is 10x what’s typical. Which I believe is a really narrow and limiting perspective. And as tempted as I am to throw this whole tainted term in the garbage, I think we can do better. I think we can come up with a definition that quits ascribing heroic attributes and starts setting goals.

Our definition of x

I propose the following:

Let x equal the sum of all gains realized by the actions of the developer, across all affected colleagues.

It’s the total network effect. If a developer (or any member of the organization, for that matter) builds, implements, or teaches something that increases 20 co-workers’ productivity by 50%, that’s a 10x effect. A real one. That exists.

And that brings our team much closer to a phenomenon we can pursue, attain, and repeat.


Originally published at revelry.co on July 14, 2015.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.