Wednesday, November 10, 2010
Interview with Rakesh Agrawal, SnapStream
For this morning's interview, we spoke with Rakesh Agrawal, CEO of Houston-based SnapStream (www.snapstream.com), which develops software which allows users to search within television shows. We caught up with Rakesh to hear about what the firm's products do, how it has transitioned into the commercial market, and where it is going next.
What are SnapStream's products used for?
Rakesh Agrawal: SnapStream is a software company based in Houston. The technology we specialize in lets people record lots and lots of TV--it's like a DVR on steroids. We make it so that people can search inside those shows. Searching inside television is not something most consumers do, but what our customer do is use it to pinpoint when something was said on a television show. For example, if you want to see all the places where Dell is mentioned on television among the recordings we've made, we can find every place that Dell shows up in a transcript, and returns that as a search result. Once that's done, you can create a clip, burn it to DVD, or email a copy of that clip to someone.
Our technology is used by lots of different people, because TV is so influential, and people are interested in what is said on TV. There are three primary places our product is used. Number one is the government, number two is education, and finally, in entertainment and news. Across those three sectors, the search is used in different ways. For government agencies, they want to know what is being said about their agency. For example, a mayor or police department might want to know what's being said about them on the news, so that if it's inaccurate, they can respond more quickly. That allows government organizations to more quickly and effectively respond to any TV coverage about themselves, and ultimately better shape public opinion.
In education, it's used as a tool to get TV into the classroom. We've made it so that teachers can pinpoint content that might be useful to them in the classroom, and pull up educational content more quickly. Say a teacher wants to put together a lesson on a topic in the news, perhaps the elections, they can search for it and pull up relevant pieces of coverage for including in classroom instruction.
Finally, it the entertainment world, it's used all over the place. Once example is the Daily Show and the Colbert Report, who are our customers. Anytime Colbert makes fun of something on the show, and pulls up that 15 second clip to make a point, the show has been using SnapStream to find those clips and integrate them into the final show. They're an around-the-click operation, up-to-the-minute, and covering what happened on the news that day.
You started as a digital video record (DVR) firm for consumers--how did you make the shift from the DVR market to enterprise some years ago?
Rakesh Agrawal: We did start out in the consumer market. We made a product called Beyond TV, which was used mostly by geeks, who would set it up in their living rooms. It enabled you to do what you can do with consumer DVRs today, which is to record TV, watch TV, and get a program guide. But, because it was built on a PC and used by early adopters, we had all sorts of cool features--things like the ability to record more than one show at a time, tape shows, and sync them to things like the iPhone and iPad. It's stuff you can't do with off the shelf DVRs you might get at Best Buy. What happened, however, is that even though we've sold a 100,000 units of the software, we found we were running out of early adopters, geeks who were willing to crack open their PCs and install software.
So we said--okay, what do we do next? We tried a bunch of little things. We created a social network for people who watch TV, Couchville, which if you look online got great reviews from people like TechCrunch. But, we built traffic but couldn't figure out how to build a real, scalable business. However, we started getting guys knocking on our doors, saying that they really liked our consumer product, because it does some unusual things--and asked us to build a version for their business. There were the guys at Saturday Night Live, there was someone at ABC Television in Burbank, and so on. So, we sat down, and thought that maybe we could go and talk to them, pay attention, and see what they're looking for. We followed half a dozen of those customers back to their offices, interviewed them, and saw how they were using our consumer product, and realized that there was a big opportunity we had been ignoring. That was exactly what we were looking for.
Why was that the case?
Rakesh Agrawal: It's a market where not lots of people are trying to solve the problem, and we bring unique technology and expertise to solve the problem. So, we went after it, and about three years ago added our TV search technology. We now have a couple of hundred customers using the product and who live inside the software on a daily basis. They rely on it as their pipeline to television. That's how we made the transition. It's not something I would have guessed a few years ago, that we would become more of a B2B company as we have become, even though we originally set out to be a consumer company. The switch has been great, and it's been good to get into business in something other than the consumer market. Now, we've built out an in-house sales team, and really changed our support process to provide enterprise support, and it's been a really successful transition.
It looks like your software is sold as an appliance?
Rakesh Agrawal: Part of what we learned with those handful of potential customers, is that most people using SnapStream and our consumer product in businesses were IT geeks. In most cases, they were spending something like 40 hours--on the low side--tinkering with the hardware to get SnapStream to work the way they wanted to use it. In order to scale this, we had to make it so it didn't require our customers to spend 40 hours to do this, so we built a turnkey appliance. Our customer can get up and running within minutes of receiving SnapStream. The appliance is turn-key, they can plug it into power, into the network, and they're ready to go.
How is the company funded and backed?
Rakesh Agrawal: We have an angel investor here in Houston, and we are mostly self funded. That angel investment goes all the way back to when we were focused on the consumer. Today, we're profitable and growing.
How big is the company, and are you all in Houston?
Rakesh Agrawal: We're a 20 person company, with one in Austin. Otherwise, we're all here in Houston.
What's next for the firm?
Rakesh Agrawal: We just got finished with a major revamp of our software. As we grew out of our consumer product, we found we were doing things on our consumer software that we built which it was never designed for. Things like Active Directory authentication and security--something you'd never ever going to do in a home. Parental control is one thing, but Active Directory is another, completely different security problem. All of that is layered on top of our consumer product. As our consumer product got long in the tooth, we started a rewrite of the core objects that comprise our software. We shipped the first version of that a couple of months ago. The key features of this new version, is it allows us to build scalable, SnapStream recording instances. For example, one box might be able to handle up to 10 channels at a time--which you might think is a lot--but there are lots of customers who need to record 30 or 40 channels at a time. The only recourse they had at the time was to buy additional SnapStream servers, and manage each independently. What we've done with our new clustering technology, is you can now use four SnapStreams talking to each other, with only one program guide to manage, and one place to search for those TV show. It really simplifies the whole process. We've also added mirror configurations for backup, in case something happens to a SnapStream. What's next for us, is we've got a couple of things we're working on, all related to the core TV search idea. Although TV search is not one that gets lots of attention, TV is very influential, and there are lots of people who still get their news from television. We're really looking to take traditional TV, and make it so people can find whatever they're looking for.