Carat has about 30 people in its data and analytics group, and it's growing even more if Mike Vitti, Carat's new "Moneyball" hire, has anything to do with it. Mr. Vitti, senior VP-group director of the data and analytics practice, joined the firm a couple of months ago after leaving his sports-data world behind. He was at the PGA Tour for four and a half years. There, he did predictive modeling to determine which players would perform the best each week. In between the PGA and Carat jobs, he spent two years entrenched in sports-sponsorship research at Repucom. Although he's just getting cozy in the media business, he talks to Ad Age about his past and how he's bringing his sports-data expertise to the media agency.
Ad Age : What made you decide to leave sports for media?
Mike Vitti: [In past roles], I dealt with both sports and media properties. But it seemed that the amount of data and what could be done with media and integrating it seemed greater in the media industry than it did in sports research. I've heard there are 56 zettabytes, which is a large amount of data, and it's available to Aegis and Carat. We have an opportunity to try to focus that and get meaningful statistics out of it. That's what attracted me.
Ad Age : What will you actually be doing that 's new?
Mr. Vitti: I want to get into computer modeling where we can run simulations and experimental plans before we execute to see what the impact would be. We have a good team of programmers and analysts. We have to make sure we're applying the right tools. There is a lot of stuff readily available. One of the most powerful and open-sourced software tools is called R. The real trick is to find expertise behind it to use it. We're also learning which data sets are available. [For example], we're trying to run things on weather data to see how it impacts the media. It's something we're working on.
Ad Age : What are some challenges you plan on facing head-on?
Mr. Vitti: Analytics groups are not involved early enough in the process. We need to get ourselves more integrated up front. One of my favorite things is designing experiments. It's a specialized area of statistics and a way to address the strengths and weaknesses of a potential team or campaign you're building. You're identifying the needs, finding those strengths and plugging holes.
Ad Age : Tell us about a time you used predictive data in sports marketing.
Mr. Vitti: I used to write articles for the PGA Tour. I was trying to do predictive performance, [anticipating] who would do well that week. I used to run these correlating statistics based on how the course performed and based on how players performed. There are different strengths for each course. One of the first years I was there, in 2005, was a year when Steve Sticker wasn't playing well. Now he's a top 10 player in the world. I looked at the game and U.S. Open course that year and predicted he'd have a really strong finish. He finished third. Nobody else had him on their radar. It was because of his putting and short game -- he was ranked near the top in a lot of areas I felt were key. The statistics identified it before anyone else.
Ad Age : How did those kinds of predictions affect player sponsorships or PGA business partners?
Mr. Vitti: One of the most interesting things was when we started a program at PGA that [enabled] universities to use the tour data to develop new statistics. We worked with MIT to develop a new putting statistic. And we worked with manufacturers on player statistics to fine-tune the equipment. They'd make a change and ask me for data to see what impact that had on the individual players.
Ad Age : How did you get into the sports-data business in the first place?
Mr. Vitti: I was doing a health-care analysis and I couldn't bring that data home on my laptop because it was protected data. So I'd download baseball data sets and write code against that to create little models. I'd do it in my spare time to test some things.
Ad Age : How will the data you collect at the media agency differ from the data you're used to working with in sports research?
Mr. Vitti: Everyone has messy data; it's not an unusual challenge. But sports data is probably the cleanest data out there. It's contained -- everyone is keeping score. All golf data comes from a single source. When you start integrating data from external sources it gets messy. In media, there are so many sources and ways to handle things differently, but it's easy to understand once you understand the nuances. Along the lines of sport, the goal is to win and it's about how you optimize a variable to win. In this case [of media], we'd optimize toward a sales or brand awareness goal by maximizing a budget to reach people you want to reach. It's the same principals.