Welcome to tvScientist Spotlight, our series focused on the people behind CTV's leading performance platform.
For this debut spotlight, meet Head of Data Science, Michael Bilow. Michael shares his expert advice for performance marketers and his vision for the future of performance TV.
(By the way, Michael is hiring! Submit your application to join his data science team at tvScientific.)
Michael is a man of many skills. When asked about what drives him professionally, it didn’t take long to discover that he's won big on game shows like Jeopardy! and The Chase. He even went on to compete in the 2015 Jeopardy! Tournament of Champions.
Amidst appearing on game shows — and after years spent in data science, machine learning research, and as a principal engineer developing semantic search techniques — Michael turned his competitive attention to adtech performance.
His creativity and competitive nature are a natural fit for optimizing advertising campaigns.
At tvScientific, Michael's data science team ensures brands' ad dollars turn into incremental outcomes. They're “performance-obsessed” — competing against the most aggressive benchmarks in the industry to fine-tune new models for optimal targeting and bidding.
Michael’s top advice for performance marketers considering CTV is to drive both activation and the upper funnel. Marketers currently using search and social channels should be asking themselves: Why are your customers searching for your business in the first place?
While CTV has become an increasingly important performance channel, it also plays a major role in driving brand awareness and audience engagement for brands. In fact, studies have shown that a targeted ad on the big screen is much more engaging to audiences than a piece of content that shows up on social or search.
Michael adds, “These days, we’re looking at TV as an activation channel. Over the next couple of years, we’re going to get a lot more information about TV as a quantitative brand channel. It’s unquestionably a better way to communicate your message — whatever it may be.”
If Michael is speaking your language, this is your chance to partner up and join his team.
And here’s a little interview tip from Michael: “Know your fundamentals. Leetcode is trash and we don't do it in data science. We're looking for technical depth — as evidence, we've been early adopters of the Zig programming language in machine learning.”
Interested in becoming a tvScientist? Apply here.