By Mindy Charski
February 8, 2013
When consumers make a purchase, turn on a cellphone, ring a call center, visit a website, click a banner ad, comment on social networks, or join a loyalty program, they’re producing valuable nuggets of data that savvy marketers can leverage to make better business decisions.
They can do this through marketing analytics, which uses software-based algorithms and statistics to derive meaning. The ability to glean actionable insights from increasingly large datasets is a boon for a discipline that, while becoming more data driven over the past 50 years, has tilted more toward art than science.
Would you like to read the full article?
“There’s always been measurement in marketing,” says Dave Fitzpatrick, director of marketing analytics at the New York consultancy Rise Analytics. “But now with the amount of data that we collect and analyze, we finally have the capabilities to thoroughly utilize that data better so the science aspect is becoming more important.”
Taking a data-driven approach to marketing can benefit the bottom line, but let’s be clear: Many marketers are not yet enjoying its bounty. The challenges come through in the findings of a 2012 study by the New York American Marketing Association (NYAMA) and Columbia University researchers: Almost two of five (39 percent) of senior corporate marketers said their own company’s data is collected too infrequently or not real-time enough. About half (51 percent) said that a lack of sharing customer data within their own organization is a barrier to effectively measuring their marketing return on investment. And 65 percent said comparing the effectiveness of marketing across different digital media is “a major challenge” for their business.
Still, 91 percent of the 253 marketers interviewed said they believe successful brands use customer data to drive marketing decisions. So while there are clear challenges, there’s also strong potential. And indeed, companies that are already using newer tools and technologies to store, integrate and analyze data are seeing the benefits of being able to more smartly target customers and prioritize spend.
Marketing Analytics Use Cases
Analytics can help marketers better understand their customer base and how well their efforts are performing—intelligence they can use to more smartly communicate to targeted market segments and allocate resources, such as deciding in which channels and how to spend advertising budgets. Brands must no longer market to every customer or prospect the same way. Instead, they can speak to individuals in a more relevant and personalized manner with customized offers that can drive sales, build loyalty, and differentiate a brand in a crowded marketplace.
For instance, to spur season pass holders to renew, the mountain resort company Vail Resorts sends direct mail that references information from the recipient’s previous ski season, including days and vertical feet skied. This data is generated by radio frequency (RF) technology on passes and lifts and is accessible to skiers through EpicMix, the company’s online and mobile application that connects with Facebook and Twitter.
“It really takes that marketing piece away from, ‘Hey it’s time to buy a product ’ and more of ‘Let’s remind you about how great an experience you had and let that experience really drive your motivation to buy the product,’” such as a season pass, says Darren Jacoby, Vail Resorts’ director, customer relationship marketing.
Marketers can also employ analytics to reveal patterns that would otherwise be hard to detect, like the one AgilOne, a cloud-based marketing analytics service, discovered for the online pet products retailer PetCareRx. After detecting a significant drop in its client’s average order value, AgilOne’s software was able to uncover a new group of customers who were buying recently introduced food products. These items have lower prices, yet this group of consumers was purchasing pet food more frequently.
“What marketing analytics was able to do is present PetCareRx with an opportunity to say, ‘You are now acquiring a new set of customers who behave differently, who are interested in different products, and figure out what you want to do about it,’” says Omer Artun, CEO of AgilOne in Mountain View, Calif.
Marketers are also turning to analytics to try to solve one of their trickiest puzzles—which marketing channels are providing the greatest return on investment. “That’s where big data comes in,” says Anil Kaul, CEO of the analytics and research firm AbsolutData in Alameda, Calif. “You need technology that brings all the data from various channels together and you’re able to build analytics that can measure the relative impact of all the different touch points you have with your customers.”
Constant Contact is among the organizations employing channel analytics to understand how new customers learn about the marketing solutions company, which uses media channels like online display, paid search and radio. But analyzing the data generated by the 35 billion emails that flow through its system annually also enables Constant Contact to offer insight to customers—many of whom are small businesses—like how to get targets to see and respond to their email messages, says Jesse Harriott, chief analytics officer at Constant Contact and co-author, with Jean-Paul Isson, of Win With Advanced Business Analytics: Creating Business Value From Your Data.
Database Marketing in the Days of ‘Mad Men’
Paul Berger, director of Bentley University’s Master of Science in Marketing Analytics program, traces the roots of marketing analytics to the 1960s with the early versions of database marketing. Two decades later, he says, bigger changes emerged: It became economically feasible to send more personalized documents—addressing recipients by name in letters, not just mailing labels, for instance; marketers became better at attributing sales to specific promotions; and companies began to focus on building the customer relationship, rather than just making the sale.
Organizations would increasingly recognize the value of transactional and demographic data, which fed into early customer relationship management (CRM) systems in the 1990s. Yet, the fragmented way many stored their structured data would ultimately restrict success in a landscape altered by the growth of online media in the 2000s and more recently, mobile media.
Many marketers found multichannel campaigns, like those that mix television and social media were even tougher to measure than single channel efforts. They also had difficulty gaining a full view of customers’ behavior across channels and predicting what those customers would do next. The situation would become even more complicated for marketers seeking to add to the mix information that didn’t fit neatly in databases, like social media mentions and call center transcripts.
Technical Advances Make Analytics More Accessible
Technological advances are making storage, integration and analysis of large datasets easier, faster and cheaper. Today open source tools like Hadoop and R that can run on commodity hardware allow users to cost effectively explore disparate types of data without first having to put them in a consistent format, for instance. “The technology and the tools are catching up to the desire of the people doing the work,” Harriott says.
The innovations translate into more opportunities for business users to search for gems. “You can go back and ask more interesting questions that you really couldn’t do before because you frankly can store almost everything that comes through your business,” Harriott says.
Meanwhile, it’s becoming easier for marketers to utilize predictive analytics, which uses statistical functions to evaluate one or more datasets to predict trends or future events. Tools to handle predictive analytics have been around for about 20 years, Harriott says, but they require technical expertise like a statistics background.
“What the industry’s looking for are those nontechnical tools that make it easy to do predictive analytics,” he says. “That’s where the industry hasn’t met its promise yet so to speak and there are a host of companies trying to solve that problem now.”
Expect to walk, not run, into marketing analytics. Here are six initial steps worth taking:
Get focused: Identify which questions you want to explore. For instance: How often should I send promotions to my high-value targets? Which of my customers are mostly likely to leave? And which offers are most profitable? Then evaluate whether your current tools can produce the answers. If not, consider the trade-offs of building in-house capabilities versus working with consultants or identifying vendors with expertise in the field.
Estimate the costs and benefits: Richard Smith, chief marketing officer of AIG Bank told a CMO Council webcast audience in December 2012 he recommends starting with a break-even analysis. After figuring out your costs, he says, project out what kind of improvement you’d need in your performance to justify the expense of improving your analytics. “Give it the reasonableness test—is that achievable?” he says. “Is it reasonable to think the investment is going to lead to benefits that will give you a return?”
Logically organize your data: Fitzpatrick of Rise Analytics says it can pay for companies to collect and store market data even before they are ready to analyze it. “Building that data infrastructure is about two-thirds of the process to getting there with marketing analytics,” he adds.
Start slow: “You don’t have to get all the way to a big data solution and use all this data you have,” says Jacoby of Vail Resorts. “Find the areas that make the most sense, use the information, test and learn, and grow from there.” In late 2012, Vail Resorts had not yet incorporated unstructured data, for example. “Our first step was, ‘Let’s just get this transactional, structured data in’ and now that we have a pretty good foundation of that, we’re starting to look at some of the unstructured social data we have,” he says.
Strengthen the CMO-CIO relationship: Marketing chiefs may have the money to bypass IT for software as a service analytics (SaaS) products and other resources, but collaborating with IT leaders could produce better results. Marketing data could be integrated with data from other areas of the organization like operations, for example, and customer information could be better protected.
Build the right team: Ideally you’ll find staffers who not only have analytical skills, but also an understanding of marketing. Given today’s shortage of analytical talent, this goal might rank among your most difficult challenges, but keep at it: These “hybrids” are out there.
Home page image from U.S. National Archives via Wikipedia