The Vision for Tealium Predict ML: In the Words of Director of Product, Dave Lucas
Tealium Predict ML, our newest paid feature for Tealium AudienceStream CDP, is generally available as of June 30, 2020. To commemorate the release, please enjoy an inside look at our Product Manager’s vision for the product. Read the full feature spotlight here to get more details on product functionality
Before I joined Tealium, I was a “data guy” for a number of brands and institutions. The common theme was that, regardless of industry or whether my team was referred to as Data Science, Decision Science, or Analytics, my charge was always to provide my colleagues in leadership and marketing with a better understanding of our users, so that we could shape better experiences for them. If we could save a few bucks along the way, that never hurt.
I wrangled many datasets, and eventually managed teams who wrangled more. We built hundreds of models using many thousands of lines of code. We got faster tools, we worked both smarter and harder— yet the backlog of requests never seemed to get much shorter, and we could never answer all the questions asked by our CEO and Marketing department.
I also know what it feels like to be on the other side— making those requests, craving those critical insights to inform your strategy, and waiting weeks or months for any progress. The pressure and frustration only compounds over time, as marketing and CX move faster and faster.
That’s why we built Tealium Predict. I don’t want to see machine learning continue to be viewed by 99% of the population as an impenetrable, pseudo-magical realm where rare experts build divining rods. Actually, that sounds cool, and we can keep up that image— but surely we can make at least some powerful strides toward making this technology more accessible and useful to the mainstream.
For the leaders, analysts, and marketers out there, we want to empower you to handle your most common use cases yourself for machine learning on customer data. You should be able to predict any user behavior, over any timeframe, to fit any campaign or experience. You shouldn’t have to write code or learn advanced statistics. The technology should come to you and fit your existing process.
And for my fellow data peeps and engineers, I want to see your backlog shorter. I want to see your time freed up from repetitive requests so that you have a bit more time to focus on the nuanced, advanced projects which make work a lot more interesting.
We’re not kidding ourselves. This is only step one of a long journey. But it’s a big step in the right direction, and I’m incredibly excited to see how it helps you.