Algorithmic technology on the web

Matt Marshall (amazing how he does it!) recently broke the news on my fund's (Dawntreader Ventures) latest investment which is still in stealth mode.  Now that the news is out, all I can say is that there is tons of data on the web, most of which is unstructured, and that the company, Peer39, has some serious algorithmic technology which can help mine that information and boost "conversion rates" dramatically for advertising as Matt succinctly describes.  Matt correctly updated that Dan Ciporin, the former CEO of Shopping.com, is not a founder but a seed investor and board member of the company.  Amiad Solomon is the founder and my two colleagues, Ned Carlson (who will be on the board) and Sang Ahn (not Dan Ahn) were leading the deal from our end.  I know - having stealth companies is a pain, but we were truly trying to keep the company under wraps until we rolled out our beta system later this year.  What we love about the model is that with some proprietary algorithms we believe we can turn just plain data into real usable information to dramatically improve the effectiveness of advertising, and that as a fund we believe there are many more opportunities on the web to take unstructured data, apply some algorithms, and turn it into real valuable information.  The other beauty is that if we execute correctly, it is an extremely scalable and capital efficient business.  More to follow and thanks to Matt for breaking the news...

Comments

[Gregg]

I really detest the term "unstructured data" because it is nonsensical. If I have random string of characters, that is unstructured. Data by definition has structure. Here is the first definition: "Factual information, especially information organized for analysis or used to reason or make decisions."

So unless they have a found a way to mine useful information from noise, the data isn't unstructured; it's not organized in a way that is useful for our purposes.