Do you remember Iron Man’s best friend, Lieutenant Colonel James Rhodes, known in film as Rhodey?
Rhodey was well educated, apparently a post-graduate from MIT. He worked as fighter pilot, who mastered rifles and pistols, but also exceptionally strong and reactive in hand-to-hand combat. He was so professional that the US Army assigned him to be a liaison with Stark Industries.
He later stole the Iron Man suit from Tony Stark. His first experience with it was not the easiest one, but showed Rhodey the immense possibilities a human could never experience otherwise.
Later, when the armor was fit to his needs and requirements, Rhodey was able to join the Avengers and become a super-hero thanks to additional speed, security, reactivity and strength he gained.
My story today is about journey of another Rhodey who is a media trader in programmatic advertising.
Since the rise of programmatic ads, the advertising market has grown considerably and quickly. The role of market traders in these real-time bidding (RTB) environment becomes even more important. Rodhey continued to decide when and where to buy the inventory for clients’ ads, constantly scanning market trends.
Only the number of campaigns and different parameters have increased, and the whole cycle has accelerated. Now Rodhey’s professionalism is defined by how quickly he finds opportunities for any given campaign in any point in its progress, depending on multiple Key Performance Indicators (KPI). The most successful traders are those who take multiple changes in the market into account and act intelligently in order to minimize negative impacts and maximize the positive ones for their clients.
Our media-trader Rodhey, like his Marvel namesake, perfectly masters the campaign configuration and monitoring, dealing with multiple campaigns from multiple clients almost simultaneously. But he is still human, and it’s more and more difficult to cope with increasing volume and complexity in RTB. It’s time to introduce him to data-scientists, who will craft the armor that fits his needs to make of him the Media Avenger.
There are three pillars required to become a Media Avenger.
The key to internet advertising is the hitting the right people with the right creative content at the right time in the right spot on the screen and on the right site/application. Even if, depending on KPI, the definition of “hitting” could be different (impressions for branding campaigns, clicks/views for performance ones, you name it), data science increases the efficiency of campaigns, whatever the definition of “hitting” is. More than just optimizing the complex process of targeting in real-time, data-science helps to adjust the price of transactions on the fly, depending on the context. With header bidding venue, it becomes a core function to get the most out of the campaign. Now the whole process of bidding doesn’t require enormous effort from our media-trader Rodhey. It’s easier for Rodhey, but also much more accurate and efficient regarding the whole RTB system.
The complexity and distributed nature of programmatic systems is a perfect soil for fraudulent traffic that provides a huge amount of fake responses. It’s easy to generate and hard to identify. It severely damages the quality of media campaigns and traditional approaches to prevent fraud, such as manually created blacklists, usually underestimate the fraud development. Data-science creates a shield that constantly detects frauds and protects the campaign, ensuring high quality traffic.
Free from tedious tasks, media-traders can now stay focused on the more interesting and complex activities, where human creativity and decision making is necessary. These tasks concern the strategies for conquering the markets.
Customer engagement is quite a complex task. On one hand, it depends on the needs of the market: strengthening customer relationships on the already acquired part of the market, gaining the new one or exploring market potential. On the other hand, social and economic context influences customer behaviour. Rodhey’s role is now to be smart and creative enough to leverage both sides, and answer questions like: “What does the rise of market share in one area and the decline in another area mean? Should I continue looking for gains in that part of the market or give up? Why is this product more popular that another? What’s next?” Reports from KPI shed some light on these questions, but to really answer them, Rodhey needs much more information to take into account. Here, the data science armor brings him hints and sometimes even clear answers , drawing from not only the patterns of customer behaviour, but also the reasons for them.
Not only does the data help to predict customer behaviour and maximally engage the customers, but the same data can impact market share building strategies.
So there we go, three pillars for Media Avengers – RTB optimisation, security and the impacts on the market share building strategies.
Those pillars lie one on the another. Like Maslow’s hierarchy, the most essential functions, like targeting and price optimisations, are on the bottom, the most creative and promising ones are on the top, with security and quality of traffic in between. We could see that data science considerably improves every stage of this hierarchy. However, to make an impact on the strategic level, we need to perfectly fit the armour to the business problems. There could be universal solutions for automation of the optimisation processes and there could be more or less universal solutions for security issues to fight fraudulent traffic. But to make the data really powerful in the hands of our trader Rodhney, we need to create personalized data-science armour for him.
Harvard Business Review research has shown that firms achieve the most significant performance improvements not by replacing humans with machines, but when humans and machines work together. So I invite you to build the artificial intelligence not to replace Rodhey, but to empower him !
” You said that nobody could possess this technology in 20 years – guess what, someone else had it yesterday! “, –