MediaGamma is at a turning point in its growth cycle. The start-up, a resident of Innovation Warehouse since 2015, has evolved and matured rapidly since joining the community and recently secured investment to take it to the next level of growth. Innovation Warehouse sat down with CEO Rael Cline to capture this milestone on the cusp of MediaGamma’s departure from the space.
What’s next for MediaGamma?
A couple of weeks ago we closed a £2 million financing round for MediaGamma. Two venture capital firms were involved: UCL Technology Fund and Parkwalk Advisors. The idea for the round is growth and expansion, so there’s a whole product development roadmap and obviously revenue targets too.
The majority of the funds at this stage will typically be used on hires (engineering, data science and sales). This round is part of our decision to move on – it’s also about being physically closer to our clients, and we’re looking to grow the team to 20-odd people, which has necessitated the move.
How would you define MediaGamma’s offering?
Fundamentally, we’re very much a pure-play data science and machine learning company focusing on online advertising, but specifically helping companies to acquire new customers more cheaply and consistently than their current providers do. Providers are typically very good engineering or media companies, but a big opportunity has opened up to add a real-time machine learning layer to the ecosystem.
Online advertising is a fascinating world. It’s an industry that generates one of the largest data sets in the world, and so it’s a fantastic avenue to showcase world-leading machine learning. We have a core expertise in doing that – we’re actually a spin-out of UCL’s Computer Science department with a specific research focus in this area, called Computational Advertising. So we can hire some of the best talent in the world, coming straight out of the university. We productise a lot of that and scale it out.
On a full-time basis we are seven people with various part-timers (interns and sponsored PhDs, for instance).
What is your client base?
It’s advertising technology companies, specifically. When you visit a website, be it on your phone via an app or via a desktop, behind the scenes there’s a real-time auction to show a user an ad. So in 100 milliseconds an auction takes place, a winner is chosen and their ad gets shown to an individual. This happens literally hundreds of billions of times every single day, and the industry itself is worth £30-40 billion and growing at double digits.
Our customers are the ones that provide the infrastructure to do the buying and selling. To all this algorithmic decision-making that takes place, we add a machine learning and data science layer. So it’s quite niche in that regard, but the actual market size is huge.
How did you tap into this niche?
Traditionally, the companies doing this are engineering companies – you’ve got a 100 millisecond window to submit a bid, so there’s significant engineering needed to help scale that globally. But there’s very little data science going on – how they make those decisions as to what price to submit, or how likely a user is to download an app for instance. Often it’s quite literally a constant bid price for an entire ad campaign, regardless of any other factors.
The finance world has been using algorithms to make similar decisions on real-time pricing for decades now and given my background in finance, I was interested to see how we could apply some of the techniques in a new market.
How did you find the experience of being at Innovation Warehouse? How has it helped you reach this milestone?
We’ve been here for about two and a half years. We started off hot-desking, then moved over to a private office, so we’ve gone through various stages.
The location is excellent – it has north, south, east and west links – and it’s close to the university, which was really important in the early stages. But above all, the community here at Innovation Warehouse, particularly for early-stage ventures, is what you’re looking for.
We’re a very technical organisation and Innovation Warehouse has good links with data science communities and engineering talent, etc. But support on the commercial side around business models; sales process; raising finance – it’s all provided through workshops, seminars or mentoring, which has been really helpful.