Boris Olij (GroupM) ‘AI helps companies achieve maximum efficiency’

Tell us about your role / position or a typical working day 

“A lot revolves around maximum efficiency and innovation if you want to get a head start in digital advertising. My job is to keep GroupM on track. As a strategist I supervise a team of experts who optimize the results for multiple clients – from local to international accounts. We do this by using programmatic trading methods and continuously improving the quality of campaigns. The use of smart technology is an integral part of this. My position therefore also focuses on stimulating the acceptance of artificial intelligence (AI) in the Netherlands, in particular machine learning.

My working day is therefore very varied and often starts early. The course for the day is often not clear until I have reviewed the team and / or client needs. For example, I need to find new CTV and audio opportunities, solve operational challenges, tackle bigger issues or attend a catch-up meeting. My schedule is constantly changing and that’s what keeps it incredibly interesting. ”

Can you briefly explain how you ended up in your current position? In particular, the path to AI innovation.

“I was lucky to be able to start a career within the GroupM family right from university, at Xaxis. I have not regretted this for a day. In the past five years, my career has not only changed in terms of function and location – from Oslo to Amsterdam – but I can also focus more on my passion: stimulating AI innovation.

I have always found AI fascinating, although I have no background in IT or software; so much so that I taught myself the coding and analytical skills for using AI. GroupM supports me in following my passion for AI and to make its use simpler and clearer. Every week I try to free up time to research new developments in areas such as reinforced machine learning and robotic process automation, which allows me to play a role in shaping the future for programmatic advertising. ”

Can you explain the concepts for non-black boxed algorithms and effective application of ML and AI in programmatic advertising to people who are not familiar with this?

“The idea behind non-black boxed algorithms is basically to get the mystery out of AI. There is a widespread perception that smart systems are too complicated to understand: data goes in, an algorithm is run in a mysterious black box and results come out. But it doesn’t have to be that way, with the right insight and good explanation, they can also be glass boxes that everyone can look into and understand how it works in there.

Program specialists must be able to clarify the less well understood parts of AI-assisted advertising. By improving their technical understanding – including how machine learning algorithms and digital attribution models work – and maintaining the pace of developments and adjustments within the programmatic ecosystem, their knowledge base is broadened. They can then use this insight to explain the programmatic operation and the results in a logical, clear and clear manner. This not only proves the value of AI, but also makes it understandable for everyone. ”

Why is AI and innovation so important to performance marketing?

“Results-driven marketing is all about achieving goals with maximum efficiency. This can be realized more effectively with the help of AI. Let’s take spending allocation as an example. If AI algorithms are used to evaluate which advertising yields the best result, budget can be targeted on the best-scoring digital media. In other words, results are optimized and waste is minimized. Or, marketers can use tools like Xaxis’ Copilot to build custom algorithms that consistently analyze data throughout the campaigns. To further predict which impressions will achieve the desired results statistically, and to automatically adjust purchasing values ​​during the process. 

The added benefit is that the processing capacity of AI is very extensive and streamlined. As a result, systems can perform continuous testing at scale and assess immense amounts of fragmented data down to the smallest fragment. In order to determine very accurately which combination of bidding timing and price works best at that moment. In addition, they feed this data back into programmatic platforms to support future decisions. ”

Do you have examples of innovation projects you have worked on?

“The best example is a custom algorithm we developed for Ford . Partly because the concept originated three years ago in a bar in Copenhagen. A colleague and I went for a drink after work and discussed how we could use machine learning to help Ford redevelop their programmatic strategy. What’s interesting is how this project has used intelligent customization to increase the precision of digital advertising and reduce costs.

At that time, the standard programmatic strategy was driven by self-generated campaign data from the purchasing platforms. Often also with ‘last-click’ conversion insights. As a result, there was little insight into the effect created by each multi-channel ad, and the purchasing algorithms were often based on a limited view of what worked well before. We hypothesized that performance could be improved by using Ford’s website and behavioral data to produce a unified view with real customer insights. After analyzing nearly ten million data points, we identified the ideal mix of factors – such as time, day and geography – to improve the likelihood of a conversion for each car model, and put custom bidding algorithms to work.

Do you work on AI innovation at Xaxis within the group and with other agency partners?

“For us, innovation isn’t just about gathering resources when problems arise, it’s also about spotting opportunities and leveraging common skills and knowledge. At an internal level, this could mean working locally within the Xaxis and GroupM team, or with a specialized team of analysts in London. This in order to increase the possibilities of our own AI optimization technology, Copilot. In the case of broader collaboration, it is within GroupM’s objectives to leverage diverse AI expertise by collaborating with multiple players, such as within WPP and the digital agency Greenhouse.

The joint project between Xaxis, Mindshare and Ford is a perfect example of the positive outcome that comes from collaboration. After revolutionizing Ford’s media buying, we are once again exploring the possibilities of adding new variables and metrics to improve the performance and efficiency of all campaigns, both locally and internationally. A similar story applies to the joint efforts with Wavemaker, Mediacom and Greenhouse.

By pooling skills, we can take on challenges, work towards scaling up, bring themes together with supply-side platforms, and much more. For example, we are currently trying to determine how different offline, online and 3rd party data sources can be aggregated and used with custom algorithms so that advertisers continue to be successful without relying on cookies. ”

What is your view on AI innovation? Where do you think the possibilities for AI lie?

“From an advertising point of view, the biggest change we can expect is likely to be found in data feeding different AI systems. AVG has put data privacy at the top of the agenda, and due to the increasing restrictions on cookies, this remains at the top. The growth of programmatic means that we now need to find new methods to gain insights while applying data securely and responsibly. Machine learning platforms, in particular, need to stop relying solely on real-time campaign performance data, leveraging a mix of 1st party customer data, offline insights, and true value conversion information.

The industry is isolating itself from traditional data, including cookies, creating room for AI. Advertisers and agencies have the ability to integrate and analyze alternative data to determine how to achieve the desired results. By combining intelligent technology and human expertise, advertisers can find a balance that gives them control over disparate data and how to use it to achieve optimal results. ”

Why is innovation so important to the industry in today’s world, especially during COVID-19?

“In these globally uncertain times, it may seem tempting to take a step back, but advertisers risk running behind the scenes. In these dynamic times, advertisers must remain at the forefront if they are to maintain their success.

At Xaxis, innovation is in our DNA, it urges us to constantly question everything and constantly adapt our technology and processes. Right now this kind of innovation is more important than ever before in order to be able to adapt to the new normal and drag advertising through the crisis. We are already seeing that the transition to virtual life is moving media channels that have not yet dealt with the digital environment at all. This is certainly not the last positive result of the current turmoil. ”

Is there someone who inspires you, that you look up to in your career?

“In my career so far and in my current role I have had the pleasure of working with many talented people, but if I had to choose one, it would be Malin Maarud, Programmatic MD of Xaxis in Norway. She probably has no idea what kind of impression she made (and still has), but Malin was an inspiration and motivation during a crucial period in my career. An exceptional manager and mentor. She helped me find my way in the programmatic landscape. The combination of her technical skills and leadership really gave me a boost in my first years at Xaxis and GroupM. Even now I follow her example in managing and empowering my teams. ”

What advice would you give to someone starting out in AI innovation?

“The most important thing is to find time to innovate, even in small steps. You may not be able to set a fixed time in your calendar for creativity and problem solving. However, that does not mean that innovation cannot be a regular part of your working day. Keeping abreast of new developments will help you identify opportunities for innovation and focus on the tools and logistics needed to make this opportunity a success.

It is also crucial to look beyond your own capabilities. Many people can be found within any company who have their own unique perspective and skills. By recognizing the value of others and learning to work with different people, you can achieve more and achieve better results for everyone. ”