Each postal rate increase sends direct marketers scrambling to find new ways to reduce or offset rising mailing costs. Experienced DMers long have relied on database marketing and analysis for ongoing cost-saving tactics to counter the escalating costs of mailing.
These tactics address the main objective all direct marketers face: selecting whom to mail and when and how often contact should be made.
When selecting whom to target, database modeling and analysis can identify customers and prospects most likely to respond to an offer. Cost savings result from selecting customers or prospects with a higher propensity to respond, thereby trimming mailing quantities while achieving the same or higher response rates.
From a cost-saving viewpoint, however, a contact strategy is even more important, determining when and how often mail should be sent to customers and prospects. Database modeling and analysis can quantitatively identify optimal times to mail. In addition, a strategy can uncover opportunities to leverage other channels, such as e-mail, to minimize mailing costs.
Yet according to one benchmark study, only 46 percent of direct marketers in the survey perform database modeling. Let’s look at the top reasons the other 54 percent give for not taking advantage of these cost-saving opportunities, and some solutions:
Poor quality data, or not enough. No matter how poor the quality or how limited customer information may be, if you can mail them, you can model them. With only the mailing address, information regarding demographics, mail-order responsiveness and lifestyle characteristics can be purchased.
From this information, database modelers can statistically identify the characteristics that distinguish customers from everyone else. In customer acquisition efforts, these models make it possible to select those lists or names within lists based on the unique demographic characteristics of the customer base.
ZIP code models are another customer acquisition modeling technique that requires only a mailing address. These analyze demographic and lifestyle characteristics that produce a model to score ZIP codes from most to least responsive. From this data, names can be selected within lists from only the most responsive, best performing ZIP codes.
When it comes to selecting in-house customers to include in a mailing, even the most basic information – such as the last time customers purchased, the amount of the purchase and the number of purchases – can be analyzed to predict response.
Finally, database modeling and analysis can build value-based segments within a customer base. By using multivariate techniques, comprehensive transactional, demographic and firmagraphic profiles of customers can be achieved.
We tried it before, and it didn’t work. What was your objective? Were staff members and outside consultants who really understand data involved in the modeling process? Was the model scored and validated? How much time elapsed between when the data were pulled for the model and when the model was implemented?
Ask the experts. Most database modeling professionals love puzzles. Find those willing to talk through the project design and results to check where something might have gone wrong.
Ask your peers or local direct marketing industry association for leads to analytical consultants who can help uncover and remedy the factors that may have produced poor results in the past.
It costs too much. That could be true when building a model for a one-time mailing. But most of the time, modeling pays for itself in just a few mailings. When researching vendors, ask to see their results and follow up with client references. Also discuss pricing options. Some outside service bureaus or list brokers offer modeling services at discounted rates because they profit from renting names or servicing databases.
Other analytical vendors may offer performance-based pricing. One example is a discounted initial fee, followed by a selection fee-per-thousand for every name selected using the model. Finally, make sure you measure and track model performance over time to determine return on investment.
We have a small customer base. Smaller companies might find it as cost-effective to mail their entire customer base as it is to mail a smaller subset. But using database modeling and analysis to optimize contact strategy can be crucial to small mailers.
For example, I happen to be a longtime, high-value customer of a small cataloger. Every month, I receive several product-specific catalogs in addition to the full catalog. I don’t even get through all the catalogs before a new batch arrives. If a cataloger analyzed my transaction history, it could see that I routinely buy from multiple product lines and categories. It could save quite a bit of money by mailing just the full product catalog to me, saving the product-specific catalogs for single-category buyers.
If you don’t use predictive models now, address any and all of these issues when researching modeling options and analytical vendors. Find experts who work with these issues daily, those who can help develop realistic expectations and attainable goals to lower mailing costs.
If you’re one of the 46 percent of direct marketers already using database marketing and analysis, look at the recency of your models and analytical findings. Refreshing or recalibrating your work can uncover new opportunities to offset current and future postage increases.
