News Stay informed about the latest enterprise technology news and product updates.

Do marketers need to be math geeks?

Technology may be getting easier to use, but some staffs don't have the expertise to execute on analytics. Peppers & Rogers takes a look at skills needed decipher customer data.

The real financial pay-off for any customer-based initiative comes from first understanding the needs and values of individual customers and then acting on that intelligence by creating different treatment strategies for different types of customers. The tool for developing the kind of actionable insight required is "analytics," through which you attempt to achieve a better, more quantitative understanding of the characteristics of different customers. Better analytics leads to better understanding, and better understanding leads to customer treatment strategies with impact. It's really that simple.

But as a discipline, analytics is inherently complex, and much more technically sophisticated than anything else a marketing person is likely to have to deal with. Not only do most marketing departments not have the competence on their staffs to do analytics well by themselves, often the company's managers aren't even confident in their ability to recognize the required technical talent when they see it, so hiring the right people can be difficult, too.

Our suggestion is that you take a deep breath, have a soul-searching moment as a company, and try to make an honest assessment of your own firm's level of expertise in this arena. You'll need to decide whether you want to assign staff to the analytics function, hire and supervise completely new talent for it, or outsource it, at least for the time being. Some specific questions might include:

  • Do you have a statistician on staff? Will you need one?

  • Will your CMO need a mathematics degree, or can she or he understand the critical issues without using statistics and econometrics?

  • How much of this would you feel comfortable contracting out?
We recently spoke to several companies, including end users and analytics vendors, to see how they are answering these questions.

Do I need to have a statistician on staff?
Although GUIs (Graphical User Interfaces) are making analytical applications easier to use and more accessible to business users such as marketers, in the end, analytic software does complex modeling and number crunching. According to Scott Van Valkenburgh, director of solutions business development for SAS, a leading analytics developer, "It's helpful to have someone on staff to tweak the models after the business users begin working with them."

However, Van Valkenburgh believes that it's "a cultural misunderstanding that you need a statistician" on staff. More and more analytical applications are available at the desktop, making data available to marketing staff through a GUI with "drag-and-drop" capabilities.

Statisticians, such as Tony LoFrumento, executive director of CRM at Morgan Stanley, might disagree. LoFrumento holds a Ph.D. in statistics and, along with two SAS-trained programmers/analysts, handles sophisticated modeling techniques in-house for 35 business units at Morgan Stanley.

"We had a game plan for determining the major goals and objectives of the organization. We then turned to the data side to find out what data was available and what we lacked in terms of applications, software capabilities and the type of capabilities we would need to run a business based on knowledge -- including predictive mining and modeling, campaign management, customer profitability and lifetime value (LTV) and performance measurement," explains LoFrumento. "We had to determine what tools were missing and what skill sets were missing and whether we had the statisticians or analysts and programmers on staff to make all of that happen."

Does my CMO need a mathematics degree?
"While it's not required [for the CMO] to have a statistical degree, it's definitely a benefit," adds SAS's Van Valkenburgh. "There's a lot of good statistical theory that can readily improve performance. To understand that, and to be able to equate marketing results to performance, is a huge asset." What's more important, he argues, is that the CMO has the management and vision skills to assemble the right staff, as well as the appropriate supporting technology.

How much of our analytics capabilities can we safely contract out?
While some executives like LoFrumento feel confident enough to handle all their analytics in-house, others are choosing to outsource their more sophisticated modeling applications.

The primary advantage of outsourcing analytics is that upfront costs are lower and you can get your program up and running quickly. But companies outsource for a variety of additional reasons, including a lack of in-house skills or inadequate numbers of staff, as well as a lack of budget for the supporting technology. It's critical to consider your total cost of ownership (TCO), which will include the cost of hardware, software licensing and maintenance; as well as the addition of qualified staff if necessary.

For example, one large retailer (who preferred not to be named) recognized that its in-house staff was completely engaged with day-to-day business processes and marketing executions, and decided to outsource very "large" analytical projects to qualified vendors. According to company officials, it was strictly a human-resources issue.

Another factor to consider when deciding whether to outsource is the likely effect of an outsourcing decision on your own firm's efforts to develop the right skill sets and capabilities. "The advantage of using analytics in-house is that it builds a core competency and brings you closer to the customer," says Van Valkenburgh. While the practice of "Analytical CRM" -- leveraging customer data to create actionable customer intelligence -- may be experimental in the beginning, companies that undertake the discipline become smarter as they go along. For example, "If you're a company like Procter & Gamble, you live and die by how you advertise and market to customers," he says. "That is a core competency and you will want to develop that competency in-house."

In other words, if your firm's business model depends heavily on your ability to differentiate customers and treat different customers differently, then even if you do choose to outsource your analytics function, you'll still want to have a plan to develop better internal analytics skills.

But the answers to the best and most important questions will only be available to you once you begin really using analytics properly, to create different treatment strategies for different customers. What behavior changes in a customer will lead to the biggest increases in value? What are the needs-based triggers of those behavior changes? How will you capture the data needed to measure and verify the actual financial outcomes? Questions such as these, which require an in-depth understanding of customer value and needs, will be our topic next month.

To read more articles like this one, visit Peppers and Rogers Group's Web site at

All materials copyright 2003 Peppers and Rogers Group - 1:1 Marketing.

Dig Deeper on Customer analytics

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

If the analysts we push into service to explain stats to us are good at their job, we don't have to add another hat to our rack of communicator, cajoler, strategist, budgeter, etc. With the right members on your team, there's going to be someone who can put spending and efforts into perspective for you.