| When Jason sent this to me as a preview, I did not think it would end up on HN. Interesting. I want to respond to a few of the comments, as I was the main protagonist in getting this idea off the ground, CEO of the company, and the business/product guy surrounded by the talented engineers who made this happen. The decision to invite the board member on to the call the day before the money was in the bank was my call. My biggest mistake of my career. I have learned from it, but it was my call and I blew it. Nothing about the situation gave us any reason to believe that we would have a problem. By the way, if anyone wants a front door key to a tier 1 VC firm, I am sure one of us still has theirs. On the topic of the business model. Let's first talk about the derivation of the idea. Having spent my career in tech and Wall St, I have seen the migration of high IQ talent into spaces where the marginal utility of IQ was trending to 0. We wanted to find an opportunity where we could be smart and do new things, and that would, theory lead to money. The base thesis here was that no one had made money in the kid protection space for a while. IQ follows money. Viola. We were right - there had been no innovation in this space in a long time, and everyone was focused on keeping porn off computers. We decided to try and focus on keeping bad guys from getting to the kids. We can all have an intellectual discussion about whether this was an overblown social issue or not. I don't feel like rehashing that here. What we did know is that we had a novel approach to solving this language problem, and we built an engine. We did not solve it with Bayesian analysis or ML. We went a different route, and the engine was very effective. Our premise was that we could sell the software to parents. We did all the things many of you would do. Built an MVP (before there was a phrase for it), talked to customers, etc. It turns out that the axiom from real estate is true here. People lie. What parents said they would pay for, and how they reacted when using the software were totally different. Many of them expected the software to come with a computer, or be paid for by someone else. They told us they would pay for it, but in reality they expected it to come from an AV provider. That never would have surfaced in talking to customers ahead of time. We then pursued a deal with McAfee. I don't know if at this point in time I can talk about that deal or not, but suffice it to say, that wasn't the golden ticket. After spending time refining the business model to get parents to pay, we (as many companies do) fell backward into our business model. It started with a fortuitous call from a large company making a massively multiplayer game for kids. It turns out many of these companies who cater to kids have rooms full of moderators for in game chat. People are very expensive. Some tech found out about our software, installed it, and found it very difficult to beat. They were intrigued enough that they asked us if they could install it inside their server farm. When we took the call, every answer from us was "yes of course" but in reality we had no idea. Credit to Tommy and Erik for figuring out in short order how to make the demo a reality. The first company figured they could reduce their people count from 600 to 60 with our software being the first pass filter. Once we figured this market opp out, we started reaching out to other companies. Long story short, we were looking at deals of $250K per year for our software. Bad timing being what it is, this was all happening around the time of the financing closing. So while Benchmark (wow, that's the first time I have said that name out loud in connection with this story in public since 2003) was moving forward on what they thought was a consumer software deal with massive growth potential, it looked like we had a business opportunity in enterprise-y software for community moderation for companies catering to minors. When Benchmark pulled out (worst kick in the stomach feeling of my life having to pass that news on to the team) we scrambled. The deals from these game companies were coming along but we weren't going to have enough runway to get them closed. The eventual buyers of the company were building similar software, and they had heard our name enough in their calls with their potential customers when being told why they were not being moved forward in the deal that they wanted to buy us. It was a little bit of good luck and timing in an otherwise crap sandwich state of play. To sum up:
1) We knew who we thought our customers were: parents
2) We were wrong about how they would act: parents didn't want to pay for this type of service, despite what they told us
3) We were wrong about the virality of the service: despite saying otherwise, parents wouldn't tell friends they were running this software because of what they perceived as potential social stigma
4) You can sometimes fall backward into your business model
5) Benchmark did us wrong, but that's life. Move on and do great things. Jason, Erik, Tommy and David are some of the most talented people with whom I have ever worked, and I rate IMSafer as the time of my life. We had a lot of fun, and built some amazing stuff (patent filing in case you are interested: http://www.google.com/patents/US20090089417)
6) Survivorship bias is a bitch, and there are many stories of smart guys not getting over the hurdle which never get told. We weren't idiots, or morons, or whatever. Sometimes things don't make it. The real question is what did you learn and how to you do something constructive with what you learned. I am happy to answer any additional questions. |