Getting to Know Your Customers

Mar 11, 2007 11:58 PM  By

Performance of a customer or prospect list is a function of many different factors: from what sources and through which channels the names were acquired; the degree to which customers have recently purchased; and how well the list is nurtured with reasonable contact frequency and kept clean from undeliverable addresses.

The best direct marketers view a list of customer names and addresses (e-mail or postal) as merely the blank canvas for painting a highly descriptive and predictive profile of future response and profitability. By building customer profiles that capture proven data elements correlated with purchase behavior, multichannel merchants can increase the performance of their lists, and substantially improve the relevance of their communications and the satisfaction of their customers.

To do so, you should focus on three types of customer information:

1. Observed behaviors. What the consumer has actually bought, by category, by SKU, with what frequency, in what season, in what size, for which members of her family, on sale or at full cost. If Web analytics can be integrated into customer data, then it is also interesting to know what she browsed, dropped into her cart and subsequently abandoned.

2. Overlay data. Long the fuel that powered better targeting and list rental in the catalog industry, demographic information (e.g. household income, family status, employment and purchase behavior) is drawing greater investments from multi-channel merchants to increase the targeting and offer relevance of their e-mail marketing.

3. Self-reported profiling. Consumers, particularly when they have a deeper bond with merchants or marketers, will tell a lot about themselves – particularly if there is an explicit promise that providing personal information will improve the shopping experience, increase the relevance of communications and offers, and generally provide them with more value.

While many columns in this space have dealt with observed behaviors and overlays, user self-profiling still remains a bit of a mystery and deserves greater attention. Why ask customers about themselves? The simplest reason is that the more you know about your customers, the better able you are to deliver offers that are more relevant and more likely to be acted upon. A more-complex rationale has to do with deeper psychological engagement: The more you give of yourself with regard to personal information, the more likely you are to remain engaged with the brand.

Personal profiling offers a great complement to observed behavior and overlay data by giving marketers tremendous flexibility to learn exactly what is relevant to their brands and product lines.

What should you ask? Begin with a basic rule. Ask too much and you risk turning people off, ask too little and you waste an incredible opportunity. We recommend a profile questionnaire of eight to ten questions that takes no longer than five minutes to complete. Specifically we recommend:

  1. The questions must be relevant to your brand. It is fine to ask “personal” questions about fit and body type if the consumer clearly sees the connection between revealing that about herself and receiving a better shopping experience.
  2. Questions should have a varied tone to them. Create a balance between direct questions about product preference (e.g., “in the past 12 months have you considered purchasing…”) and questions that take on consumers’ attitudes (e.g., “With regard to shopping for consumer electronics, I consider myself to be (a) highly knowledgeable…”)
  3. Stick to single-response or multiple-answer, highly structured questions. While it may be tempting to have opened ended questions where consumers can add their deepest thoughts, you can only target on explicit answers to explicit questions.

How does one implement a personal profile? Here are the key steps:

  • Draft your questions, keeping in mind the tips above. But remember, it’s not advisable to change questions in the future as it is best to maintain consistency of both the questions and responses for the purposes of targeting.
  • Develop internally or outsource the “programming and hosting” of the survey instrument. A number of practical, easy-to-program solutions exist that allow marketers to design simple surveys, linked to Websites and e-mails, that nicely capture data in the formats that database teams can read and upload.
  • Determine and execute a compelling user experience. Begin by asking, “Why should I answer these questions?” Companies with existing loyalty solutions offer points in exchange for their members’ time; without that “give-to-get” mechanism, you must really emphasize and deliver a better customer experience going forward.
  • Reach customers at the right time. Lower your expectations for e-mail campaigns. While you will get your best customers – and that’s great – it will be tough to penetrate into your next best and average customers – the segments where better information can really move the needle. Focus instead on initial registration, when you have their attention; and post transaction tools such as confirmation pages, confirmation e-mails, and shipping confirmations.

Asking your customers what they think, how they purchase, and what they want from you has direct and powerful impact on both response rate and customer satisfaction. Marketers who target based on profile response can increase response rates by five to ten times over less targeted, less relevant, less engaged campaigns because communications are better suited and because consumers are more involved.

Finally, customer profiling can be tested and rolled out over time. Pick a small sample of your customer base, selectively invite them to update their profiles and apply the insights to a handful of marketing campaigns. Then, take what you have learned and invest in getting as many customers to self profile as you can.

David Rosen is senior vice president of Loyalty Lab, a San Francisco-based developer of customer loyalty programs for the retail industry. He can be reached at david@loyaltylab.com.