Mikko Kotila, Founder & CEO, STATSIT Discusses the Evolution of Their Proposition & Regional Differences in APAC.

Can you give an overview of the STATSIT proposition and how you have evolved in the data-driven display-buying space?

STATSIT started on the back of our previous company, which was providing automation and workflow management solutions for some of the biggest SEM players in Scandinavia. We used the money we made to start STATSIT in 2008. The first 18 months were funded out of our own pockets, so we were literally free to do whatever we wanted.

We had a small business development team in the US with ex-employees of RevCube and X+1 and our tech team was in APAC. We ended up developing a stack that included web analytics, SEM, social and also had a DMP coming. Then sometime in 2009, when we were just about to start integrating with DoubleClick, we almost ended up running out of money (and out of business).

We stepped back and looked at the market at the time, and decided that social analytics was the one of our solutions with the fastest revenue potential. At that moment, we had two weeks of runway left and zero cash coming in. We went to some of our closer relationships from previous lives, and made them an offer they couldn't refuse.

With the business back afloat, we ended up innovating in the intersection of social data and advertising effectiveness until 2011, when the first trading desk popped up in APAC. For us that was like coming home.

Would you define your proposition as an independent trading desk? Do you work with agencies and marketers?

We work with agencies and clients. You can say that we are an independent trading house, but not in the sense of an agency trading desk, as we see them here in the region. I think many of them still operate in a rather old-fashioned way. I believe it comes down to culture and the way they are resourced.

Incentives are another thing that the big agency bosses should look at. Nothing helps planners and suits to shift spend into programmatic quite like bonuses do. Until this shift happens, clients might benefit from being extra careful with the online media portion of their media agency's recommendations.

In comparison to the big media agencies and their desks, we're small, lean and totally independent. We've mostly funded ourselves, so we have a bit more flexibility than those companies who are part of a group, or have institutional funding. This also forces us to have extra focus on developing competitive solutions.

Has STATSIT developed its own targeting solution for clients? Can you give some insight into how it works? Is it a semantic solution? Does it compare to something like Peer39?

We have brand safety, contextual targeting and quality-control solutions that work on-site and at page level. Our contextual solution is a little different from what you see in the market typically.

Basically, the segment (e.g. automotive) comes alive for each client and starts to evolve in a Darwinian style survival of the fittest. We don't believe in lossless computing much. Instead, we try to look for inspiration and intelligence from nature, like the swarming intelligence of the Weaver Ants.

We're looking into the most aggressive and resilient life forms found on our planet. Like bacteria and cancer. In the zero-sum game of media trading, it's all about staying power and the ability to beat anything that gets in your way. Nothing does that quite like bacteria or cancer.

We're big believers in small data, lossy storage and fuzzy logic. I think ad tech is a great starting place for a much larger revolution in computing.

You've invented a ‘quality score’ for the inventory across which you trade. How do you score impressions and high-quality long-tail sites? Is it a mix of algo and human verification?

It's a bag of tricks that we've mastered over the years. We don't get into details with how it works. Every real trading house needs a black box, that sits deep in ours.

What it does for the customer, is that it floats the cream on top, so to speak. I like to think that we are the best in identifying relevant high-quality mid- to long-tail sites, at a very large scale. Furthermore, the fact that none of our data comes through publisher relationships helps us to stay independent in this sense too. Our customers know that we have no incentives towards big publishers, for example.

As well as using this for your own trading, you've also integrated it as a targeting option within other DSPs. With whom are you currently integrated?

We've signed deals with Brandscreen, Turn and ADZ, so far, and are currently going through integration and testing.

From both trader and technology solutions, you have a good view of the market in APAC. How is the market developing, particularly Japan and SE Asia? Will we see massive adoption of data-driven media buying in 2013?

Japan is always its own thing, and already has some local players operating at an impressive scale. For example, MicroAds may very well end up being a force to reckon with outside of Japan as well. They have more customers, more spend under management and more third-party data than the whole of SEA, Australasia, HK and Taiwan all combined.

SEA is developing poorly. There is no regional exchange. The trading desks are underdeveloped and are not getting the resources they need. Campaign budgets are in the low five-digit range. Worst of all, there are less than 10 local startups in the space in the whole of SEA.

During the last year, we've seen serious interest and investment in our ecosystem by global players like Turn, and that's great for the whole industry and market in SEA. I hope we see more players and more money this year. Also, my hat is off to you guys at ExchangeWire for the APAC section and running ATS in our neck of the woods.

In short, I think 2012 was ‘ready’, 2013 is ‘set’ and 2014 will be ‘go’. At least for those of us who will still be around.