The Robert H. Smith School of Business’s Center for Complexity in Business held its second-annual Digital Marketing Analytics Roundtable Aug. 1, 2013 at the Smith Suite in the Ronald Reagan Building in Washington, D.C. The roundtable was held in collaboration with Teradata. Smith marketing faculty members Bill Rand, the director of CCB, and Roland Rust, the executive director of CCB, led the event.
Because it was the second time the event was held, it was “time to dig deeper,” Rand said. He noted that the marketing community needs to create an “analytics culture” to figure out the science behind marketing, instead of relying on “hunches and intuitions.”
Much of the roundtable focused on the increased number of marketing channels available and how marketers can use marketing analytics to determine which of those channels reach customers most effectively.
The roundtable was divided into three components: track, analyze and optimize.
The “track” panel included Simon Zhang, business analytics director for LinkedIn, Matt Comstock, Razorfish’s vice president of business intelligence, and Smith Marketing Department Chair P.K. Kannan.
Zhang discussed how LinkedIn has used marketing analytics to keep LinkedIn’s members active on the website, noting that the company has increased member engagement significantly through the use of email.
By increasing LinkedIn’s member growth and engagement, the company is able to achieve a critical mass of data, which enables them to create products and services that are relevant to LinkedIn’s user base, which enables the company to achieve additional membership growth and engagement, Zhang said.
Kannan’s presentation focused on research he did with Smith PhD student Alice Li on attribution modeling and optimal resource allocation. Costs and effectiveness can vary widely among paid search, organic search, social media, referral sites, direct visits, email and display ads, he said.
He discussed how a website’s customer conversions may mostly come from the same channel, but that doesn’t account for customers who make multiple touches to the website from multiple channels. Therefore, specific sources of customers may be more or less effective than they initially seem.
The “analyze” panel included Cathy Burrows, director of strategic initiatives and infrastructure for the Royal Bank of Canada; David Abbott, senior director of online marketing for Home Depot; Eric Rivard, chief operating officer of Cerulium Corporation; and Wendy Moe, academic director of Smith’s MS in Marketing Analytics program. Rebecca Bucnis, Teradata’s global program director of Consumer 360°, moderated the panel.
Marketing departments literally have to do more with less these days, as budgets are shrinking, but there are more channels for marketing, Bucnis said.
Abbott noted that this means every dollar in a marketing budget needs to be analyzed in order to determine what elements are, and aren’t, driving success. “If you’re targeting everyone, you’re targeting no one,” he said.
Smith’s new Marketing Analytics graduate program, for which the first cohort arrives in September, is aimed at addressing the lack of talent in this area, Moe said.
Rivard, who was on the panel to discuss Cerulium’s client Verizon, discussed the company’s unique situation of operating in a saturated market and the challenges that presents in trying to attract new customers. Verizon is “drowning in data,” but needed to figure out how to use the data to target customers using the channels on which the customers want to be targeted, he said.
The optimize panel included Craig Oldham, vice president of digital engagement for the American Red Cross, and Professor Rand. Oldham focused on how the Red Cross uses various channels to achieve its mission more effectively, including how to better allocate its resources in response to natural disasters.
A meeting between Oldham and Rand at the 2012 roundtable led to Rand and his students studying how the Red Cross can determine the most effective channels for seeking donations.
Because about 1 percent of tweets published to Twitter are geo-tagged with the sender’s location, Rand was able to compare tweets from Hurricane Sandy in 2012 and Hurricane Irene in 2011. The two hurricanes followed similar tracks along the U.S. east coast, but Sandy was much more destructive and required a significantly larger response from the Red Cross than Irene.
Rand found differences between the tweets for the two hurricanes that could help the Red Cross increase donations. He noted that, during Irene, there was more discussion on Twitter of subjects besides the hurricane than during Sandy. He also found that most of the social media traffic related to Irene occurred before and during the hurricane, whereas the peak of traffic related to Sandy came after the storm.
- Peter Haldis, MBA Candidate 2014, Office of Marketing Communications