Eno River Analytics

Eno River Analytics was my one-man shop for applied econometrics and statistical programming in Durham, North Carolina. It was my second solo venture. Unlike the first, which happened back in college and never made a sale, this one paid the bills between August 17, 2007 and September 10, 2010.

My typical client ran a small to mid-size consulting shop. They had done graduate work in economics and had been doing their own Stata programming ever since. They occasionally outsourced some of that work, to cover spikes of uncertain timing and magnitude in their workload. I only had to address two important concerns: one was code compatibility; another was over-reliance on code written by other people. Other things equal, Eno River Analytics would stay in the black as long as I could demonstrate an ability to follow general directions, set up quickly and proceed with little supervision, producing transparent code of high quality, on time. By other things I mean chiefly a market for this sort of service, delivered by a guy working out of his home office.

I usually wrote code directly on on my clients' machines. I built it over a VPN connection, and stored it in the client's own Subversion repository. The client would own the source code, all the revisions, and the documentation. By the end of the engagement, we both learned a few things we could use later. I became familiar with the client's general requirements, so subsequent engagements would have lower start-up costs. The client would have picked up a few new tools that would make both their own work and future outsourcing easier.

I could have worked in SAS or R in a pinch, but I had used and occasionally taught Stata since the fall of 2000, so I preferred it strongly. I had a good time with it; the market for custom Stata programming is not big, but it will keep a small shop plenty busy.