Opinion: Keeping the Luster on Direct Mail
While direct marketing is extremely effective in reaching targets, frequently it is the first area to be cut from marketing budgets. Why? Because it is the most measurable.
Here's what happens: Management places what's quantifiable under the microscope and says, "Let's improve the numbers or reduce the dollars, possibly moving them someplace else."
Perhaps this is one reason why spending in direct mail is only expected to grow by 5 percent in '06 when spending in online advertising is expected to grow almost 30 percent and public relations by 10 percent. Arguably, this is an unfair predicament for marketing's shining star.
The Case for Using Incremental Modeling in Direct Mail
One of the most important strengths of direct mail has been its ability to measure specific customer reactions not always discernable via other channels: the number of items purchased, leads generated, products activated and spend levels.
Today, this alone is not enough. What's also needed is an assessment of the true effectiveness of the campaign, one that evaluates whether the cost of the creative effort, premium, distribution, etc. was worth the investment.
Traditional modeling tools can help to meet this requirement. But incremental modeling can make a dramatic difference in improving marketing productivity and return on investment (ROI). Incremental modeling does this by reducing costs and generating higher response rates, thus increasing the ROI. Let's look at how incremental modeling might work in the case of a bank as an example, with the thought that a similar approach would work for your specific market or for a client in another industry.
Historically, banks have used some form of statistical or standard modeling to identify the best prospects or customers to reach for a specific product or service. While this provided a way for the bank to rank order and find the best responders, it did not discriminate between the two different types of customers:
Those who would act regardless of whether a campaign was undertaken.
Those who required a campaign to "nudge" them to respond.
Incremental modeling enables marketing professionals to focus the campaign on the customer segment that needs additional encouragement to carry out a desired behavior. Through the use of direct mail incremental modeling, it's possible to more effectively reach the right universe of customers, cut costs directed to "natural" responders, and improve response levels.
The 'Nudge': How Incremental Modeling Works
Let's assume a bank's senior management wants to encourage more of their customers to activate their new debit card. To address this situation, the marketing department implements a direct mail incremental modeling process as a way to identify those people who need an incentive or "nudge" to act.
Generally speaking, steps in the process would include three phases: testing, development of the model and application of the incremental model.
In the first phase, testing, the course of action is to:
· Set up and execute a test mailing, including a holdout "control" group.
· Evaluate the results at the end of the promotion period. This includes measuring the difference between the group that received the mailing compared with a control group that was randomly pulled out of the mailing, but met all of the same criteria prior to the mailing.
· Next, begin developing the incremental model. The primary objectives in this part of the process are to improve customer segmentation and flag key prospects. The action items would be to:
· Review information the bank has on file regarding those debit card customers from the time when the names were selected for the test mailing, including data elements such as the date of direct debit authorization (DDA), age of cardholder, current activity for that DDA account, etc.
· Incorporate additional pieces of demographic data into the analysis, ranging from household income to the presence of children.
· Create a model that identifies those characteristics that are most predictive of accounts that require a "nudge" (such as an incentive received through direct mail) to activate.
Last, finalize and apply the incremental model.
· To define segments for future "scoring" activity, cutoffs are established in equal 10 percent increments, or deciles.
· These results are used to assign response values to a particular customer segment based on a calculated score.
· An economic analysis is initiated to determine the return on investment of mailing to each of the deciles.
· The completed analysis and results are used to determine the best mailing quantity and customer mix to maximize results.
· From this process, then, a targeted campaign is initiated and results are evaluated.
Comparing Standard vs. Incremental Modeling Results
Let's take a closer look at hypothetical example, comparing the results of a direct mail campaign using standard and incremental modeling initiatives. We'll compare a standard and incremental model, assuming the bank selects the four top deciles and mails to 400,000 customers as part of a standard model and mails to only 200,000 customers under the incremental model.
Keep in mind, using the standard model, the marketer is mailing to a larger group of customers, with no focus on "natural" responders (those who would activate without the mailing). In addition, to compare the standard model results to an incremental model, revenue that would have been generated whether the mailing was done or not (i.e. revenue from unmailed customers) is deducted.
Assuming a cost of $0.36 per piece mailed, the incremental model produces an ROI of 58 percent, four times higher than the 13 percent ROI for the standard model.
If each model consists of at least six mailings, the total dollar return of the incremental approach spread across each of these mailings would be nearly $140,000 -- clearly, a number most bank marketers (and their CFOs) would find very attractive.
As you know, direct mail has long been one of the most reliable, effective and, yes, measurable of marketing channels. Regardless of what industry you serve, with a little science like incremental modeling, it's possible to greatly improve your direct mail ROI, the primary measurement most organizations employ to determine how well (or poorly) a particular project performs.
Tom Salutz is Chief Technical Officer of DataCo Solutions in Arlington Heights, Illinois. He can be reached at firstname.lastname@example.org.