Using Your Operational System as a Marketing Database: Just One of the Seven Deadly Sins

Companies fail at database marketing for a variety of reasons, says Jeff Hassemer, director of product strategy at the data management solutions division of marketing solutions provider Abacus.

During a Webinar held Feb. 23 by MULTICHANNEL MERCHANT and “Direct” magazines, Hassemer cited several of the more common reasons:

  • A lack of focus on what matters and the resulting “project scope creep.” According to Hassemer, “Often companies attempt too much at once with the database and allow multiple groups within a company to submit requirements.”
  • Insufficient data access and data accuracy—or conversely, overly stringent requirements. “Sometimes the data just isn’t there,” Hassemer said. “Don’t try for perfection.”
  • Insufficient time to develop, deploy, and realize the value of the data.
  • Failure to plan for and execute strategies that drive ROI. “Create a plan to implement and stick to it,” Hassemer advised. “More importantly, create a three-, six-, nine- and 12-month plan for how you will use the database once implemented – and stick to it.”

And then there are what Hassemer considers the seven deadly sins of database marketing, the first of which is using your operational system as a marketing database.

“It’s the poor man’s marketing database,” Hassemer said, adding that this is most often done when budgets are tight or IT resources are limited. “This approach rarely works for anyone, because often the information is just not available for the marketer to make effective decisions.”

Hassemer detailed why using your operational system as a marketing database can be deadly:

  1. Operational systems are generally designed to put data in, not take them out. There are few places where a marketer can go in these systems to get one single view of the customer or even understand all of his transactions in a single place.
  2. Because a marketer cannot get a solid view of the customer, he is making decisions that simply are not correct. Hassemer used this example for multichannel merchants: “I call in from work and I do not have my catalog with me. The call center rep misspells my name and creates a new account for me. Am I a new customer? No, but a marketer may select me as new and send me into a completely incorrect contact strategy because of it. Am I a valuable customer? Yes, but the marketer won’t know that because they are going to select me as new.”
  3. The operational system does not allow for aggregate views of the customer.
  4. The operating system doesn’t allow for reporting or its reporting conflicts with input into the database. “Reporting is generally operationally oriented and doesn’t provide real insight into how customers shop,” Hassemer said. “Can you act on the reports that you get from the operational system? My experience tells me no, because the clients that build marketing databases generally get a completely new view of the purchase behavior by doing so.”

Fortunately there are three ways of avoiding this sin:

  1. Build a specific marketing database, or data mart, from a data warehouse. A data warehouse is a central repository where one cleanses, consolidates, and stores data for use in separate functions; a data mart is a repository designed for a specific function or purpose. A marketing database should be built as a mart from a clean source such as a data warehouse.
  2. Update your marketing database regularly. “Update frequency is determined solely by your company’s ability to analyze and make decisions upon the data,” Hassemer said. “The frequency of your update cycle depends on how often you will change directions based on the data. In general, a weekly update is enough to keep the data fresh enough to keep up with today’s fast-paced environment, yet infrequent enough to keep everyone sane and [allow you to] do some decent analysis.”
  3. Design the data mart for access to the data you will need, and make sure reports can be run quickly. “No one wants to wait for answers, and an inefficient database will not get used,” Hassemer said. “Limit linking in reports as much as possible, and design the aggregations around the type of information that you want to get to.”

For the rest of Hassemer’s seven deadly sins of database marketing, log on to https://multichannelmerchant.com/events/webinars/seven_sins/ for a free replay of the Feb. 23 Webinar. And if you have questions for Hassemer, you can send them to him directly at the “Ask the Experts” section of MULTICHANNEL MERCHANT’s “Website: http://expert.multichannelmerchant.com/database_marketing/