How we put Yieldex together

Every company has a “how we got started” story. For example, here’s the NetGravity startup story. This is the story of how I got involved in Yieldex.

First, a little background (bear with me – this is relevant). After we sold NetGravity, and I took some time off, I was thinking about a career change: get a PhD and become a computer science professor. I’ve always loved to lecture (a trait shared by many VCs, which is not necessarily a good thing), and I thought I might enjoy the pure intellectual stimulation of academia.

I began by looking for an interesting topic, because I got some advice from professor friends that most PhD students spend the first couple years just trying to pick their thesis topic. One hard problem we had worked on at NetGravity was the “overlapping inventory” problem, so I wrote a short paper on the topic to present to the HPTS Workshop, a fairly exclusive enclave of the top database academics in the world, including Jim Gray, Ed Codd, and Michael Stonebraker. Mostly I wanted to see if they could tell me if the problem had been solved before. I got generally good feedback that it looked like an interesting and unsolved problem.

For a variety of reasons, I ended up going into VC instead, and put the problem on the back burner.

Doug Cosman was one of the key employees at MatchLogic, another early internet advertising startup that was bought by Excite. Doug’s work after Excite led to the key insights that underlie the Yieldex technology, and he raised some angel funding to pursue developing those ideas. Doug was searching patents and other sources for prior art during his patent application when he came across my paper on the HPTS web site, which by then was ranked highly by Google. In a fairly random coincidence, one of Doug’s angel investors at Sequel Ventures, Chris Scoggins, is a friend of mine, and put Doug in touch with me directly.

I have to admit, I was pretty skeptical that some guy in Boulder had solved a problem that had pretty much stumped the industry for the better part of 10 years – classic Silicon Valley arrogance. I put off talking to him for some months, but fortunately he was persistent. Finally he came out to visit, and in a four-hour session with the whiteboard, convinced me he had a novel and workable solution.

Learning about his solution re-ignited my own interest in the problem and the space. Of course, it didn’t hurt that DoubleClick had just been acquired for $3b and aQuantive for $6b, and the space was hot in general. The more phone calls and meetings I had, the more clear it became to me that this was still a huge problem for the industry, and that nobody had solved it well. Also, for me personally, the process of doing diligence and getting back in touch with a number of people reminded me that I had a lot of personal credibility and expertise in this area. Very quickly I made the personal transition from wanting to invest to wanting to be involved.

Fortunately, Doug and his angels had already talked about hiring a CEO in the Bay Area, so it was fairly straightforward from that point on.

The Woodside guys have been great to work with, they have been incredibly supportive and helpful. They invested in the Series A, and I’m still very much a part of the Woodside Fund family, now as a Venture Partner and a portfolio company CEO instead of a Managing Director.

2 Responses to “How we put Yieldex together”

  1. Mark Mouse Says:
  2. Todd Vernon Says:

    Tom, enjoyed the presentation at VCIR. At Lijit we are just on the early curve of figuring out the complexity of ad scheduling but your presentation resonated with me. Would be fun to get together and discuss.

Leave a Reply