I found the whole thing interesting just because they’re sort of a mix of models in one company that focuses on a less-popular segment (bare metal). They also try to be a vertical in a very innovated and cost-competitive space. That gets a lot of businesses killed for getting too far behind or too expensive. Yet, what nearly killed them was not being ready for piles of customers coming their way. Advises to be ready to quickly scale up for several times your current business but that’s startup 101. It seems they put quality before growth which was a mistake given advantages of being market leader. Can always improve quality later.
Really weird, but interesting, case study.
EDIT to add: Maybe a good example of Gabriel’s Worse is Better where those following that approach won while Online was trying to do The Right Thing in terms of design or ops quality.
devops might be a better fit here than web.
Also, thanks for the datacenter porn–I wish they had more shots of that converted fallout bunker. :)