Conversation

Cedric Chin
@ejames_c
Something I’ve been toying with: the software company equivalent of capital efficiency isn’t tech efficiencies but go-to-market efficiencies. In other words: the cost edge comes from relatively cheaper acquisition not relatively cheaper servers.
Quote Tweet
Dan Luu
@danluu
I don't agree with the reasoning that Kubernetes will be forgotten because it reduces productivity. If you look at the most successful post-Twitter tech companies, backend effectiveness and efficiency is just not a serious driver of success. It seems to barely matter at all.
Show this thread
Tweetbot for iΟS
1
Quote Tweet
25
Likes
Cedric Chin
@ejames_c
Replying to
@ejames_c
I don’t 100% believe this, to be clear. Empowering a small engineering team to shave off cloud costs seems perfectly reasonable in some situations. But maybe not the highest order bit.
1
4
Cedric Chin
@ejames_c
(Of course, minutes after I tweet this, my brain generates multiple counter-examples. Fraud at 2000s-era PayPal, server spend at some infra startups today — e.g. every new free-tier user spins up a large instance. So obvious exceptions exist).
1
3
Riley! 🐀
@riley_stews
Replying to
@ejames_c
Labor costs far more than infra for all but the most compute heavy software. Being quick is the sign of effective labor, typically.
nonsense pronouncer 🇺🇦
@the_pebcak
Replying to
@ejames_c
I think you’re on to something. There’s a lower bound of operational cost below which no net new companies are enabled to succeed. Like if it costs $1000 a month to start a company vs $100, really how many more successful ones would exist?