Sramana Mitra: So roughly speaking, it is end of 2007. How much venture capital did you raise at that point and how long did it take you to get the first product out of the door?
Eric Burns: We had an extreme advantage here, which is that we have the opportunity to research that market to understand what classroom capture required under the CMU umbrella for several years.
Sramana Mitra: You did all the research and validation work before the company was founded. It was all happening on a different clock – not on the venture clock.
Eric Burns: Exactly. By the time we incorporated, we released a V.1 within a month of officially starting. This is because all we had to do was button up what we had used at CMU and make it distributable. Keep in mind that it was an academic research project. It was not coded like enterprise software. The very first thing we had to do was to figure out which pieces we could save. We saved some of the C++ and the client application code. We ended up using those for a couple of years after starting but the server was written on an open source stack on Pearl. It was really not fit to go into production beyond the CMU hooked up to all of the monitors. We rebuilt the entire thing on Windows server and on a dot net stack. It was about six months before we were out with a rebuilt server.
Sramana Mitra: With the first version that you released – let’s call it the minimum viable product – what did you do in terms of customers? How many customers did you ship that to? Were they paying customers?
Eric Burns: Initially, we had CMU and University of Pittsburgh. Pittsburgh had paid. The University of Pittsburgh had been a Panopto customer since before there was Panopto. After it had been successfully implemented at University of Pittsburgh, we started looking at schools in the Pittsburgh area, which we exhausted pretty quickly. Then we moved to the Northeast and Southeast. It was very much the traditional startup story of hustling and getting on planes and making promises that you couldn’t quite meet unless things went well.
Over the first year, it was, as you might expect, a bit slow going but we got in a couple of key accounts like LSU Law, North Carolina Central University Law School. We were, in a sense, not selling a product. We were selling a technical team and a leadership team and the promise that we would somehow do better than the systems that had been brought to market. A lot of the people that bought early on were buying a dream. I’m sure that’s a common theme in your stories.
Sramana Mitra: This is a question that strikes me. I went to MIT and the whole OpenCourseWare effort from MIT has been active for a long time. Does MIT do its own technology for lecture capture then?
Eric Burns: I believe it does. MIT is a place where if it wants some technology and nothing is within arm’s reach, they can just build it like CMU and Stanford. By the way, Stanford has built their own lecture capture system. It’s interesting you mentioned OpenCourseWare because that’s a bit different from the space that we’re in. The product, in a sense, is the raw materials to build a course teaching whatever it is that you’re trying to teach. That’s a long term, for-the-good-of-the-academic-world kind of project. We’re a bit different. We’re a tools company. We were building a technology product that solved a niche need and our product was the tech – not the content. The content always belongs to our customers.
Sramana Mitra: I understand. For OpenCourseWare to be possible, you have to do lecture capture because all these MIT lectures are captured and put on this OpenCourseWare distribution platform.
Eric Burns: Exactly. The question you’re raising is one of scale. If you’re putting a handful of classes on a system like OpenCourseWare or in a more modern equivalent like Coursera, edX, you’re favoring, to some extent, quality over quantity, which is why you see these numbers thrown out - $50,000 to $100,000 per MOOC is the rule of thumb. That’s because they’re exquisitely produced. There are people in studios and blue screens. They’ve got AV technicians. On the other hand, you are not going to reach 80,000 hours of capture a year or where University of Arizona is – 3,000 hours a week – unless you’re prepared to accept a slightly lower level of quality and also unless you had a system that can do almost everything automatically. The difference with that kind of video is that that was a deliberate attempt versus we have a mandate to record every class that’s taught in every classroom for the next four years because this is our contract with our students. That’s a different kind of problem.
Sramana Mitra: The purpose is different. I was just curious about what these big universities who had been doing online video lectures for a while were using. My guess was they were using homegrown technology.
Eric Burns: By the way, one thing that we’ve observed is that the top two universities are laggards on this front. They tend to be lagging behind, particularly the second and third tier institutions in terms of their adoption of this technology. If you look at our customer list, you’ll find huge number of schools that you may have heard of but you don’t see in news all the time. They are years into a strategy of capturing every class in every classroom. It’s an interesting observation.