Where Industry Meets Innovation

  • Contact Us
  • sign in Sign In
  • Sign in with certificate
mit campus


Search News

  • View All
  • ILP News
  • MIT Research News
  • MIT Sloan Management Review
  • Technology Review
  • Startup Exchange

ILP Institute Insider

August 29, 2016

How Going to Mars Improved Marketing

DataXu applies its MIT-developed technology to understand customer behavior

Steve Calechman

The end goal of marketing hasn’t changed much over the years: drive actions by getting a brand’s message out to consumers in the most effective way possible. The challenge is that the landscape keeps evolving. There are more ways than ever to reach consumers; campaigns can be executed in weeks now rather than months; and making adjustments in real-time is not only possible but also mandatory. Budgets are forever tight, and with the number of potential channels available, it’s easy to spend on ones that are not effective. Using analytics becomes a key ingredient. This is the landscape in which DataXu operates. The Boston company applies its MIT-developed technology to understand customer behavior, and uses those details to help marketers target their inventory and experience the biggest returns on their investments.

Sandro Catanzaro

In from outer space
The origins of DataXu go back to an MIT Department of Aeronautics and Astronautics project for NASA. The goal in 2004 was determining the best and safest way to get humans to Mars and back, says Sandro Catanzaro, DataXu’s co-founder and senior vice president of analytics and innovation. The research being done was focused on eliminating unproductive decisions up front. With NASA, the technology revealed that out of the 30 billion possible combinations of rockets, orbits, engines, capsules and mission lengths, only 1,162 were actually viable, and those were analyzed in depth. “A far more feasible number of combinations than 30 billion,” Catanzaro says.

As one of the project researchers, Catanzaro says that he saw possible marketing applications for the technology, and DataXu was launched in Boston in 2007. Since then, it has opened up 10 more offices across the United States and Europe in 2011 while adding Singapore and Australia to the mix in 2015. Today the company has 17 locations and over 360 employees.

The company’s platform gathers data about individual opportunities to reach consumers. When a person loads a webpage, in less than 20 milliseconds, DataXu’s automated advanced analytics can calculate the probability that an ad will drive traffic to the advertiser’s website, resulting in a “success forecast” for a specific campaign, Catanzaro says. The result is that marketers can avoid wasting money on unproductive messaging. “It’s like poker,” he says. “If you know the true chances of the cards in your hand, you would pay no more and no less than their actual value.”

This forecast, or “probability function”, then learns over time and can further break down findings by demographics, knowing, if advertising a car, how men, 18-25 years old, may care more about the engine, and women, 45-54, might focus on safety features.

The final piece is the ability to measure when consumers have been exposed to enough advertising. Data maps these saturation points, and marketers can readjust investments, says Catanzaro, adding that the overall result can mean a 10-15 percent increase in efficiency and sales volume. As an example, after running on DataXu’s technology, a British telecom company discovered that it had the choice to either save $14 million pounds or create 100,000 more customers per year.

This kind of innovation at the company is driven by a symbiotic relationship with clients, such as Ford, Lexus and General Mills. “Our competitive edge has come from listening to their challenges and coming up with solutions,” Catanzaro says. “Pushing the boundaries of what the technology can do benefits all of our clients, and provides a first-mover advantage to those with the best ideas.”

Science that works in the marketplace
Catanzaro came to MIT in 2004 to study at the Sloan School of Management. His background was in mechanical engineering; his hope was to expand his business skills. While studying management, Catanzaro ended up getting a second master’s degree in aeronautics and astronautics and experienced what the Institute produces on a daily basis: ideas that are technically sound and also make business sense. “At MIT, the combination between the best management and the best science is what makes the environment so truly, truly special,” he says.

That combination doesn’t guarantee immediate acceptance. Catanzaro says that his company’s typical client is a large business-to-consumer (B2C) company looking to acquire more customers or build more quality market share. Organizational resistance to analytics is somewhat typical: embracing a new platform requires a shift in an organization’s approach and thinking. There’s a need to not only learn the technology, but also to change operations by adopting full automation in order to leverage and experience new efficiencies.

When it’s embraced, though, companies end up becoming extremely nimble and closer to their customers, he says. By combining multiple data streams into insights, automated analytics allows for decisions that are fact-based and timely, realizing consumers’ needs even before consumers are fully aware. “Analytics does not replace wisdom, but supports it, allowing it to evolve faster,” Catanzaro says.

Senior executives who are initially hesitant often get fully onboard once they see how DataXu’s technology improves sales volumes and drives smarter, more strategic investments. Catanzaro says that he’s seen the transformation. The attitude is no longer about merely liking a technology or wanting to upgrade processes. It becomes a kind of must-have. “Companies not taking advantage of data are losing a critical edge,” he says. “Data is not a new trend. It’s an imperative in order to remain competitive.”