Theresa LaMontagne calls herself a "data native." Having seen many facets of the data industry throughout her career, she's deserving of the title. Ms. LaMontagne got her start at decades-old consumer-research firm Simmons where she helped reposition its magazine-measurement product. From there she moved to BBDO.
"I wanted to learn how all this data was used," she said.
Later on, her work at Yahoo as director of research for sales gave her a strong sense of the types of digital data available to marketers, and how to evaluate quality. Today, following stints working on data teams at Carat and OMD, Ms. LaMontagne leads a robust data division at MEC. Three years ago she began building out the firm's newly formed Analytics and Insight practice as managing partner, senior practice lead.
Her goal in crafting the practice was to ensure that data are viewed holistically. Who says digital data should be used only to inform digital efforts?
"I believed that was the right way to go from an agency perspective," she said. "People think that [media channels] are so different, and they're really not." In truth, the main differences lie in the jargon, she said. "It's like buzzword bingo … but it's still media."
Ad Age: What do you wish marketers would understand about what people who work in data analytics do?
Theresa LaMontagne: Data actionability requires close integration with the business. Wikipedia describes the "infinite monkey theorem" as stating that "a monkey hitting keys at random on a typewriter keyboard for an infinite amount of time will almost surely type a given text, such as the complete works of William Shakespeare." Good data leads to great insights.
However, while we need processing power and code to harness data, we must also marry an understanding of the questions at hand with knowledge about how the available data can be used to answer them. The full potential of data cannot be realized unless the people sitting on the data are grounded and trained in the business and have a handle on the questions being asked. In short, machines don't develop insights, people do, so don't put your data people in the corner.
Ad Age: Sure, the data folks probably don't get as much respect as they deserve sometimes. But they're not perfect, either. What do you think is problematic about data people as they navigate the world of marketing?
Ms. LaMontagne: The integration of data science professionals into the marketing industry is still a fairly recent phenomenon. The historic separation of the disciplines has created the equivalent of a tower of babel with each discipline developing its own language and world view.
In media there is never a clean petri dish and often optimizations that appear to be mathematically sound do not align with the realities of the business or the marketplace, which hinders the actionability of the analytics.
Additionally, media planning has always been a mix of art and science. Perhaps Einstein was the first media researcher when he said, "Not everything that can be counted counts, and not everything that counts can be counted." While it's critical to organize, parse and sort data into databases, dashboards and reports, it's equally important that we focus on the right metrics and analyze them in the broader context of everything we know about the consumer and the marketplace.
Ad Age: Lots of digital data vendors talk about how their targeting models are better than others, or provide more reach than others. With your background on the sell side, I know you are especially skeptical of promises like these. What are some issues surrounding audience targeting data that affect your work at MEC and how do you vet data vendors to separate the true look-alikes from the look-nothing-alikes?
Ms. LaMontagne: The continued rapid growth of tablets and addressable TV is resulting in the digitization of other mediums, which will spread audience targeting across screens. As we scale audience planning across clients we need to start to better understand why something works or doesn't work which will require a deeper understanding of the underlying data sources and the predictive models used. Unfortunately, buyers are currently limited in their ability to reliably gauge the quality of the underlying data, specifically:
Quality and accuracy of the underlying targeting source data
Definition of the data sources used including the size and age of the underlying data set
Audited assurances and standards to ensure consumer privacy is not being violated
Full disclosure of methods including the validity of the underlying predictive models
As an industry we've worked together since 1964 to ensure the accuracy of the estimates produced by the various ratings vendors such as Nielsen and Arbitron, but thus far we've done little to extend those standards to this "new" trading currency.
Ad Age: What's a common misconception about big data you'd like to dispel?
Ms. LaMontagne: The first: that all data is good data. The single biggest driver of actionability is the underlying quality of the data. If it is inaccurate or incomplete it doesn't matter how much of it you have, or how quickly you get it, it just won't reliably tell you anything. The second is around the definition of "big data." It's becoming the new buzzword but most marketers don't have "big data" -- they just have a large amount of data. The third misconception is that putting all of your data in one place is the same thing as having a data strategy. Databases must be designed with the end use in mind. A thoughtful and proper database design is a key to all future data mining.
Ad Age: What's the coolest or strangest type of data set you've ever worked with and why?
Ms. LaMontagne: I have always been fascinated with marketing-performance data. Creating a successful advertising campaign is dependent upon the right mix of ingredients -- both media and message attributes. No two campaigns are the same and success is dependent upon getting the optimal combination of six variables: medium/placement, the frequency of exposure, the timing of the campaign, the audience/reach, the competitive environment and of course the message itself. Each of these components influences how well the advertising works or doesn't work and forgetting any one of them can lead to problems.
Many people boil all of these down to one measure: ad response. While it's an important element to look at, what we are really trying to understand is how that ad response -- either passive or active -- impacts customer equity. Successful advertising (i.e. media and message) interacts with an audience to cost effectively influence attitudes and behaviors (in both the short and long term). It's about determining the right mix and value of audience response, understanding the natural limitations of both the underlying data and the media marketplace. That's what I call cool.
Ad Age: What educational fields of study and professional backgrounds help develop the best data people?
Ms. LaMontagne: Great data people can come from a variety of backgrounds. What truly separates "great" data people from "average" data people are the softer skills such as creativity, curiosity, problem solving and strategic thinking. If you listen closely, no two questions posed by clients are exactly alike and there will never be a perfect data set. Therefore, even in this age of data abundance, analysts need to be creative on how they solve for gaps or flaws in the data and leverage what they have to answer the questions at hand.
Lastly, the best data analysts for marketing are really great at understanding the consumer experience. They have the ability to step into the shoes of the consumer and link themselves in a personal way to the behaviors they observe. In doing so, they are better at uncovering the meaning behind the behaviors to apply it to marketing.