Look Within: Your House File May be Rich With Data

Apr 14, 2008 11:41 PM  By

Mailers faced a challenging year in 2007. And unfortunately, due to the economy, a sluggish housing market and the upcoming election, things are off to a slow start in 2008. Continuing list fatigue and declining list universes are causing prospecting response rates to plummet. Mailers must be more strategic about reducing expenses by mailing smarter.

But never fear. Some of the best names to mail are right in your own backyard – older buyers, requestors and retail shoppers. One would be surprised how often mailers overlook these important segments.

Recently, our firm performed a best practices study for a $50mm, multichannel marketing company. In analyzing this mailer’s house file contacts and optimization, we uncovered major opportunities. To our surprise, we discovered that only 30% of their available buyer file universe was being mailed.

Needless to say, there was ample opportunity to reactivate these records using a variety of methods, such as marketing promotions, scoring and co-operative database optimization. In the case of this mailer, we elected to change their recency, frequency and monetary (RFM) scoring system to a micro-level view while also scoring the house file with multiple hits.

After the merge, we incorporated Abacus’ Dynamic Batch Optimization (DBO) product to enhance selectability by scoring records based on product affinity. It should be noted that each of the co-operative databases uses similar first-class reactivation tools—each mailer should determine which works best for their particular offer.

This DBO process allows a mailer to control batches or types of names that have the greatest affinity for a mailer’s category offer. In the case of this mailer, we created three batch files—buyers, requestors and retail records – and each batch was scored with the following four rounds of product affinity:

Round 1 – Targeted product affinity: This is the core product control model. This category is the most specific and highly targeted, relative to the mailer’s customers’ purchasing habits.

Round 2 – Giftware: This category is a noteworthy, but less targeted category, relative to the mailer’s customers’ purchasing habits.

Round 3 – Lifestyle: For this mailer, a lifestyle product select is an important variable.

Round 4 – Balance: These records make up the balance of names not scored or selected in the first three rounds. Consequently, their scoring takes no specific product category affinity into account – only pure recency, frequency and monetary habits are considered.

Creating specific product affinity rounds allows a mailer to select deeper into specific product affinity behavior scores. In more cases then not, particularly among niche mailers, product affinity behavior is more important than general RFM behavior. Therefore we can mail more names with greater affinity, but with lower RFM characteristics, and still receive very strong results.

You might wonder why buyers, requestors and retail shoppers are separated into unique batches. We do this for two main reasons. First, these segments will have different aggregate response variables that, in turn, allow us to make different selection decisions. Second, inactive buyers naturally have more mail order activity at higher RFM scores then non-buyers. They would, if we optimized all records in one batch, force requestor and retail segments into lower optimization segments. This means those requestors and retail shoppers will likely not be selected.

By separating these types of names into unique batches, we can strategically select types of names based on various product affinity and RFM.

Michelle Houston is a vice president at San Rafael, CA-based catalog consultancy Lenser.