The goal of customer acquisition analytics is to use data to fine-tune and improve direct marketing selections. But how does a marketer know when data elements will be truly predictive, learn what characteristics will be most powerful for customer acquisition needs and determine whether the model gains that are indicated will produce winning results?
Marketing goals vary by industry, and the use of data analytic services varies drastically within each. This article reviews the fundraising, catalog, publishing and finance sectors to show how the proper combination of data and analysis helps marketers find new customers.
Fundraising acquisition analytics. Analytic efforts use past mailing experience to bring strong future results. Because affinity is so essential to list success, straight selects provide the majority of prospect volume for fundraisers. Lists are selected from market segments defined by interest coupled with demographics appropriate to the offer.
Fundraising mailers source most of their donations, though not all, from seniors who possess much disposable income. Certain fundraising causes, such as those related to politics, the environment and health, appeal to broader populations demographically linked by a passion for specific causes.
Quality data coupled with regression modeling helps expand fundraisers’ list options. Regression models balance and weight many additional lifestyle and demographic variables to improve marginal performance. Regression models and other analytics can leverage the predictive power of data elements related to lifestyle interests, purchases, channel and offer preferences, demographics, area-level characteristics and so forth.
Data also can be aggregated to geographic areas to help provide lift. All things being equal, a segment of marginally performing names will improve when poor-responding ZIP codes or carrier routes are removed. For example, the majority of top prospects for a health cause typically reside in areas where the fundraiser’s existing donors live (birds of a feather flock together).
Marginally performing straight selects from compiled lists can be made profitable with the aid of analytics and quality data that target the right prospects for a fundraising offer.
Catalog acquisition analytics. Catalog mailers derive much of their prospect mail volume from other catalog lists. Large catalog cooperative databases contain transaction recency and affinity information or category spending data that easily qualifies prospect names, with or without modeling.
To combat attrition and maintain a robust, profitable active file, catalogers’ acquisition mail plans may allocate a segment to rental lists and tolerate lower performance thresholds to obtain new, unique customers.
This incremental volume often is obtained through multivariate regression models that balance affinity and transaction data along with overlays of lifestyles and demographics. Common sources of new names include publisher files, compiled files and self-reported product registration databases.
Though names from these sources will be marginal performers compared with top-tier catalog co-op names, rental lists appended with transaction, lifestyle and demographic data permit the catalog marketer to dial in new prospect volume when needed. Also, catalogers can improve results just as fundraisers do by using ZIP code filters to weed out poor geographies from a list or mailing.
Publishing acquisition analytics. Publishers often use promotional upfront offers, such as premiums, trial issues or sweepstakes, and look for a combination of upfront response and back-end pay-up in their acquisition efforts. Some publishers mail hard offers that sell on value. Top-tier lists include affinity and near-affinity hotlines, plus channel specifics such as direct mail sourcing.
Analytic profiling can demonstrate differences between paid and unpaid responders. Indicators of general financial status – such as high income, home ownership and presence of a credit card – frequently lift back-end performance. One typically successful segmentation approach involves eliminating pockets of clearly unprofitable names, such as non-credit cardholders, then creating one or more regression models to identify names that have potential.
Another suggested practice: model gross responders when conversion rates are comfortably above 50 percent, and net responders when offers are promotional and deliver less than 50 percent pay-up. Alternatively, to qualify prospects from large marginal lists, two-step modeling for upfront response and profit can be used.
Data indicating channel and purchase preference is highly valued in modeling. Transaction preference data – such as past use of direct mail, telemarketing or the Internet – or data indicating a general interest in shopping, will bolster a model’s predictive strength and value by promoting good direct mail prospects and removing consumers who do not respond to mail offers.
Financial acquisition analytics. Credit card and insurance acquisition marketers seek large volumes of names to mail frequently. For many financial mailers, potential revenue from new customers is very high compared with the cost of mailing. Consumers characteristically buy financial products and services during periods of life change. Marketing messages and top-tier lists focus on life-stage triggers such as moving, marriage, having a child or retiring.
Financial marketers’ appetite for list volume typically greatly exceeds the availability of new mover, new birth and marriage lists on the market. To fill mail plans, financial marketers leverage analytics that use demographic, geographic and aggregate financial data in models to identify likely responders. Strategies can differ, however, in instances where credit prescreening is required.
Customer acquisition regression models used by financial mailers employ large databases with demographic variables such as age, income, home ownership and aggregate financial data. Using two-step models, marketers can dial-in more response or higher approval rates as their business goals require.