Simon’s Blog

Well the rumour is that Google is going to buy Riya (the face photo searching company), for something like $40M. Is it true? No idea. If it is, it’s pretty interesting. It’s been obvious for a long time that extracting semantic meaning from unstructured data is an important area; and I’ve often wondered why companies like Google, Yahoo! and Microsoft haven’t done more in this area themselves. Aside from being an important business area, this kind of stuff is really fun to work on - which, incidently, equals very happy developers.

As an aside, my other company, Ash Biotech (a life sciences company), is currently taking things a step further from content-based searching of static images - we’re developing some pretty damn cool systems for super-detailed, super-accurate content-based searching of video. Over the years, I’ve been lucky enough to work at the cutting edge of a range of really interesting software areas, such as: content-based image searching; natural language understanding; 3-D graphics; robotics; distributed systems/SOA; high-performance/super-computing; and machine learning/AI (genetic algorithms, simulated annealing, neural nets, SVMs etc.). And I can safely say that content-based video searching is the most fun of the lot!

But I digress. Why would Google buying Riya for $40M be so interesting? Well because as far as I can see (and I don’t know the Riya folks, or their investors, so I don’t have good info here), that would be valuing the company mainly for its technology. Recently, when these kind of acquisitions have been made, the valuations were based on some formula based on revenue multiples, user numbers and/or strategic value (e.g. Skype, Flickr etc.). In other words, the valuations have been nothing to do with technology.

It has been a good five years since anyone actually valued technology platforms per se in any serious way. As wrong/unfair as it might seem, during the past few years, the usual value assigned to technology platforms in trade-sales, VC financings and IPOs has been as close to zero as makes no difference. I wonder if this might signal the start of a return to people actually valuing technology again?