Dimitri Maex is managing director of the New York office of OgilvyOne Worldwide, which is the customer engagement arm of the advertising agency Ogilvy & Mather. He’s also the co-author of the 2012 book Sexy Little Numbers: How to Grow Your Business Using the Data You Already Have. He helps clients uncover and act on insights. Maex spoke with Data Informed about roles big data and analytics are playing in a field not traditionally known for having number-crunching prowess.
Data Informed: How has analytics in advertising changed since 1998, when you joined Ogilvy in Brussels as an analyst?
Dmitri Maex: First of all it’s become really hot. I started as a lowly analyst in a corner in the Brussels office and nobody was really interested in what I was doing. I had to really sell myself. I don’t have that problem anymore and neither do any of the analysts here. All of a sudden analytics people are sort of the rock stars of the agency world.
I think the clients’ focus on accountability has been part of what has driven that. They expect everything to be measurable and more data driven. And the explosion of data has helped that as well. It’s a lot cheaper to get your hands on the data and it’s starting to become a lot easier to analyze the data as well.
On the technology side, on the data manipulation, on the analytics side, on the deployment of the findings, a lot has happened that takes the manual work out of the equation. Things that used to take me days or weeks now only take an hour, which definitely changes the way you work.
Another thing that happened was back in the day, you would gauge the size of a company’s analytical capabilities by the number of statisticians they had on their payroll. More recently it’s no longer about the number of people. It’s more about the processing power and the algorithms.
What are some questions big data can answer for ad agencies?
Maex: “Who do I talk to?” is a very important one. Direct marketers have known that forever—that the most important driver of performance is actually targeting and finding the right people. Once you know who to talk to, what is it really that will interest your target audience? What do they feel, what do they want, what do they need, and how can you use that insight to deliver something of value to them? Where do you find them?
That’s where data, especially digital data, over the last couple of years, has really taken off. It’s about a lot more than just finding people geographically. It’s finding them in real time at the right moment in time so you intercept them almost at the point of purchase.
Then, “Does [the campaign] work?” and that’s the measurement piece. “What do you do to get better?”—that’s [about] testing. And then there’s “How much do I spend on marketing?” and “Where do I spend it?”—not just in terms of media but also broader marketing objectives like “Do I spend it on awareness versus creating demand?” And, “Where do I spend it geographically and by segments?”
You’ve said there’s an opportunity for data to make marketing a more experimental process. How so?
Maex: Before digital, you put a campaign in market and you needed to wait weeks if not months for results to come in. By the time you found out it didn’t work, it was too late to do anything about it. In digital you can try things quickly and see if they stick or not. You can use digital as a lab and if things work, you can bring them to national media or to broadcast media, which I think is an interesting way of doing things.
Are creatives using big data to help them craft ad campaigns?
Maex: If you think about the creative side, coming up with the idea is one part of it and then there’s the fine-tuning the execution of the idea. I think the latter can be a lot more data driven.
In the media side what we do a lot are these creative meta-analyses and we crunch huge volumes of data and look at, in addition to placement and format, what are some of the creative variables that could drive performance? Is it, for example, pictures versus text or is it a certain color or certain types of pictures of people versus product? We’ve done a lot of work in that area and you can use multivariate analytics to identify which creative variables are driving performance.
That will help you in optimizing the execution, but it doesn’t necessarily help you come up with the big idea. That part, in my opinion, is going to need to involve a [strategic] planner or a couple planners who are very data literate and very good at distilling the single insight or a couple of insights that come out of the data . . . For example, we’re using search data in quite a strategic way where you look at what people are searching for and that often has tremendous, very interesting insights that can spark creative ideas.
What are some of the challenges agencies need to overcome to successfully deploy analytics?
Maex: There are multiple challenges. If you think about the whole value chain, there’s getting the data and having it in a place where it can be accessed quickly. That’s what a lot of agencies are working on right now and agencies are generating a lot of data themselves for the first time in many years, often data the clients don’t even have, and that’s more on the media side. The biggest challenge there is how do you take all that data and integrate that and how do you interrogate that data in a way that’s relevant for the agency’s end product? On the media side, that’s often a lot easier because you’re making decisions on how much to spend, where to spend it, and data can help you answer those questions. On the creative side it’s often harder because the end product is not as tangible and it’s harder for data to inform the creative process.
On the talent side, most agencies need to get fresh talent in the door, people who are open to working with data. It hasn’t been the industry people go to because they like numbers. It’s quite the contrary. But now it’s becoming very much a data-driven industry, especially on the media side, so you see a huge change in the talent pool. Most agencies are making those moves, but all of a sudden, from a talent side, they’re competing against industries they’ve never competed with before, so people in finance, people in IT, people with the likes of Google or IBM. The additional challenge agencies have is they don’t just need to find people who know mathematics or statistics. They need to find people who also have creative sensibilities and know how to work in a marketing environment, which is a really hard profile to find.
The last one is process. Agencies need to find a way to embed data and analytics in the core of what an agency does . . . The real challenge is how do you not just have [analytics] as a sexy shiny object or as an afterthought but how do you embed it in the way the whole company works?
What’s ahead for data and advertising?
Maex: One of the most exciting things that is starting to kick off are connected devices. The more we’re going to be interacting with smart fridges, with smart cars—all those devices are going to generate more data, so that is exciting in itself because you’ll get of-the-moment data of people interacting with the devices. These devices can also become channels, so you can start to communicate with your customers through the devices.
If you take the example of a fridge and if you’re a Unilever, for example, or Procter & Gamble, who traditionally have never had access to the end-customer information, they can get information about what people are using through these devices, either they come up with these devices themselves or they do partnerships with companies that do. It opens up a whole world of possibility that we don’t think of today.