An Open Letter To Tony Hsieh (CEO of Zappos)
messagefortony.com
messagefortony.com
What if they find their architecture doesn't scale and they really can't handle it -- is it worth the risk of going under because you spent all your savings on servers and employees to handle this company-changing client, only to lose them immediately when things don't work out right away?
I guess that's why startups seek funding, eh? I'm not sure what I'd do if a Zappos wanted to sign up with one of my apps, honestly. Spend the money scaling up quickly for them, not knowing with certainty that I'd be able to keep ahead of their usage, or just turn them down?
Also in my experience it has been a waste of time to think about scaling. Big companies do not simply look at your website and tell the lackey engineer to integrate your product into their production product and pray for it to work. There are going to be many phases of adoption which your ability to scale will be tested.
Scaling is a nice problem to have, worry about making something people want first.
I'd love to know, for example, how a company like MixPanel or GinzaMetrics goes from handling a few million data points per customer, to perhaps billions. It seems like making the leap from helping other startups to helping Disney could take not just exponentially more resources, but a whole different architecture.