Sramana: In the first few quarters of 2010, you released a minimum viable product by outsourcing to India and hiring some freelancers. Can you tell us a bit more about that process?
Sramana: Did you use a company in India?
Graham Cooke: We used oDesk. We hired an individual freelancer in India.
Sramana: On oDesk you built a team of freelancers that augmented Emre’s work?
Graham Cooke: Exactly. I think we ultimately got rid of that code pretty quickly. We hired out first full-time engineer that June, and we started to re-write everything ourselves. We were able to make the technology evolve in the direction we wanted to go.
Sramana: That all led up to your first 10 customers. Can you tell me a bit more about that mix of customers?
Graham Cooke: They were a mix of retailers and services businesses. Most of them were quite small. Our model at the time was to provide the model as a project-based service. We intended to put the technology in place to gather data and build knowledge. We would then deliver a report and get paid between £15,000 and £25,000. We quickly realized that it would be hard to make that business work. We realized we would be much better off if we could get our customers to pay us through an annuity licensed model.
Sramana: In the summer, you decided it was time to raise financing. What was the thought process there?
Graham Cooke: We were not sure if we wanted to go with a group of angels or a VC. We decided that we would pursue angle funding because we thought that we were too early for the VCs. We put our word out to the network and got recommended to a really interesting angel investor. He had done a few tech and property investments. We thought it was a great fit, and we ended up raising $1.3 million from that family.
Sramana: What were you able to accomplish with that investment?
Graham Cooke: We were able to get our own office. We hired a lot more engineers, and we really started to develop technology to be smarter. We realized that the valuable part was collecting the data. Instead of plugging into web analytics systems, we decided we wanted to become an analytics system. We really wanted to focus on visitor behavior which meant understanding much more about what a user does on an individual level. We started to build our own tracking and data collection technologies. That required a bit of investment on our part. We were able to leverage AWS and we were able to build some cool stuff that has scaled to where we are today. Today we handle 15,000 queries per second, which is gigabytes of data per hour.