Tom Bowman (pictured below) is certainly no stranger to the advertising world, with eight years under his belt at BBC Advertising and a decade at Microsoft prior to that. In an interesting move, Bowman has hung up his publisher hat, with his recent appointment as CEO of the newly founded demand-side platform, Intelligent Optimisations (IO). Here, Bowman speaks exclusively to ExchangeWire about his new role, his new company, and the direction this new market entrant is heading.

ExchangeWire: You spent eight years at the BBC, working on the publisher side. What attracted you to ad tech and how are you finding the change?

Tom Bowman: The BBC is a great global brand, and I loved working there. I was hired to take the international web service commercial and I learned a huge amount in my time there, and largely achieved what I had in my head that I wanted to do.

However, I’ve always been interested in new and emerging technology, and in the art and mystery of advertising.

The automation of online advertising, and the development of ad exchanges, matches the growth in inventory and is entirely logical. And yet, the true potential of programmatic has yet to be realised. Software intelligence is the key way to add value for advertisers in their approach to the exchange.

When I was introduced to IO, I immediately realised that what they had was part of the future of advertising – the ability to efficiently reach unknown audiences with superbly relevant advertising.

IO was co-founded by the ex-chief IT architect at the Met Office. What compelled him to transition into the ad tech space and build a DSP?

Faster [big] data. Weather data is updated every six hours, whereas in real-time advertising the response has to be, well, in real time! Weather data is normally six-dimensional, which is tricky, but at 40+ dimensions, the IO data has so many more aspects that it presents incredible challenges to model, process, or even conceptualise.

This was a technical challenge, combining scale and intelligence within an unprecedented cycle – mere fractions of a second. Alex [Nicholson]’s approach and vision has attracted a brilliant team of data scientists and platform specialists who have created a genuinely powerful machine-learning-driven prospecting platform.

Given that the DSP market is already well-established, with many successful players, why have IO built a DSP from scratch?

Many DSPs predate what has only been possible and economically viable in the last two to three years. Building from the ground up enables new efficiencies and new capabilities and gives us a platform, agnostic by device, which can move the market forward and move with the market; and, in that sense, is genuinely future-proofed.

Aside from being a proprietary technology, what is unique about IO? What enables it to competently compete against more established DSPs?

It’s not so much that the technology is proprietary, it’s what it does for customers. We are able to deliver real competitive advantage by overcoming the limitations of pre-determining your target audience and looking at all online auctions across the exchanges. IO searches 100% of the UK internet, giving marketers an enviable reach of over 45 million prospects across 90 million online devices. By searching on such a huge scale, we find entirely new audiences in unexplored places – delivering the purely incremental customers that other DSPs cannot reach.

Whether that is buying a car, booking a trip, or renewing your home insurance, our advanced application of machine learning provides a step change in prospecting quality for marketers.

In addition to this, our technology processes and analyses an exceptional level of data that brings invaluable insight into customers beyond an advertiser’s core demographic. We analyse this information, and provide dynamic detailed reporting, that highlights real-time opportunities to finetune the campaign for maximum effectiveness. These actionable insights can also be applied to other media, proactively increasing performance across a brand’s entire advertising campaign.

This is where the real value for advertisers is created – the perfect way to reach new customers without having to make assumptions.

You are eschewing the traditional audience targeting methods in favour of a broader approach to prospecting. How receptive are marketers to this?

It has been very positive. We’re extremely pleased with the testimonials we’ve received from existing clients; and delighted within our short life span to have been nominated for this year’s Performance Marketing Awards in the Most Effective Use of Programmatic category.

Our aim is to deliver the promise of programmatic so that marketers come back to use IO again and again. By prospecting further and more effectively than any other DSP, we substantially increase our clients' reach to identify the largest possible number of qualified prospects. This reinvigorates the sales pipeline, fuels campaign performance, and increases return on investment.

We are a performance business, it is all about results and delivering a real competitive advantage.

What has market reaction so far been to the results being yielded by IO?

Looking at data across some of the campaigns we have run over the last six months, we saw average improvements in Cost Per Action / Conversion of 63% by end of month one, increasing to as much as 90% by end of month three, which really showcases what Machine Learning can offer to programmatic.

The market clearly recognises this effect on CPA targets. By injecting machine learning into programmatic platforms, in some cases we literally managed to reverse plummeting product revenues and even produce a small year-on-year uplift.

The longer a campaign runs, the more improvements we see in conversions. This is due to the constant self-optimisation machine learning offers to campaigns: the longer they last, the more data is being crunched, the more we learn, and the bigger the uplift in performance.

IO is headquartered in Plymouth, with a satellite office in London. What are your expansion plans?

We have big plans for 2016. On top of bringing machine learning to programmatic and introducing our True Prospecting solution to market, we also expect to see significant geographic expansion this year. The company has grown from five employees in 2013, to 50 today, and is expected to double in size by the end of the year. After solidifying our position in the UK market, naturally expansion to other key markets is on the agenda. Watch this space.