Given Tumblr's explosive popularity, there's no way that they aren't a giant ball of stress right now. It's certainly nice to have $40 million, but that's of li

Given Tumblr's explosive popularity, there's no way that they aren't a giant ball of stress right now. It's certainly nice to have $40 million, but that's of little use in the short term: As Fred Brooks famously pointed out, adding more programmers to a late project makes it later. So the people who understand Tumblr's systems are almost certainly working nights and weekends and going slowly insane. And there's no relief for them: If Tumblr hires more staff, their current engineers will still have to deal with all the crises and train the new people. It's the CEO's job to deal with disappointed and angry customers. But it's also the CEO's job to deal with programmers who are working 70 hour weeks and who are just about ready to snap and move to a commune. Startups can be full of bad craziness. In a perfect world, customers would talk to the CEO and the customer support team, and not try to directly contact the engineering team. That's not to say that Tumblr handled this especially well. But again, it's not an easy situation.