author of mirdata.report 
Why start a Data Consulting Business in 2016?
Recently I left my best job thus far, the job at Looker, to found a consulting company. Naturally there were a lot of factors at play, but Looker’s gravitational pull on me was due to the technology itself. Looker is one of those technologies that everyone wants to build on top of because:
- It is cool.
- It is permanent.
Since the early 2000s, there have been plenty of technology stacks that my peers would get excited about. For me the earliest such tech was the Mozilla Foundation’s open browser. I would develop an extension, and within just a few months of my uploading it on the Mozilla site, my invention would already have thousands of downloads. You could just feel the momentum behind Mozilla.
Of course, there were some technology stacks that never lived to see their audiences grow. I have generally bad memories of these, but if you ask any developer, they all have a few such bad bets behind them. It is a very unpleasant feeling to make one of those bad bets. Not only are you investing more upfront time to overcome the steep learning curve, but you are also not getting the desired results. The sponsoring company just goes down and takes the tech with it, albeit slowly.
As a consultant, you are also making bets on technologies. To consult for companies in tech, you generally have to focus either on something very safe and legitimate (e.g. Oracle, Cisco) or something brand new and exciting. The former world is full of people who dedicate their entire lives to specializing in one technology. My father is such a consultant, with a couple dozen Cisco certificates to his name. I have the utmost respect for what he does, but having spent an entire year at age 16 to acquire one such certificate, I knew immediately that this competition was lost before it started. And ever since, I’ve tried to focus on the latter: new and exciting.
When I joined Looker as an employee two years ago, its tech was precisely that: new and exciting. To me and to many, it clearly had (1) the cool component. As I left Looker recently to start a consulting practice, I now see the safety in (2) its permanence.
Naturally there will be many vendors, tools, and applications that I will have to support as a consultant. Right now my sweet spot is in HDFS, AWS Data Pipeline (or anything AWS related), BigQuery/Redshift, and Looker for a data stack. The competition among vendors is fierce at every level of this stack. Luckily, I have some insider information, as financial analysts would call it, to feel confident about at least one technology in this stack ;)
This picture is here purely because no self-respecting data-driven technologist would publish a post without a picture