Want to transform customer insight into multichannel success? All you have to do is take an analytics-driven approach. But that’s typically easier said — and envisioned — than done.
You first have to overcome several challenges: identifying and defining critical data elements, removing data silos, recognizing data tied to a unique customer, accounting for missing data, creating integrated customer data views, coordinating testing and marketing plans across media channels, and tying “best customer” response to media mixes. But the dividends of this discipline bear fruit.
How do you derive and make use of such insight? How does a better understanding of customer behavior translate into more sophisticated targeting, lead nurturing, and a more robust customer-focused marketing strategy?
The first step is to understand and respect the role that customer information itself plays, and the absolute necessity to extract, discern, and then analyze unique customer data.
INTEGRATE YOUR VIEW OF CUSTOMER DATA
Regardless of what you call it — customer relationship management (CRM), database marketing, or customer optimization — being “customer centric” means connecting with customers in a way that is meaningful to them and, in turn, delivers value back to the business. So the first order of business is an integrated view of customer data through database platforms that link disparate systems, and a data quality regimen that enables clean data through effective data integration.
True integration interweaves all functionality within and across business areas to support more effective customer interactions and thereby increase customer value. In most organizations, the marketing department — in its zeal to link and integrate such data — most often serves to create this centralized view of “customer truth.”
For relevant customer interactions with the retail brand, all areas of the organization — marketing, customer service, finance, point of sale, inventory, etc. — must operate from what essentially is a shared customer data repository. Media channels, too — direct mail, e-mail and mobile marketing, outbound calling, point-of-sale “clientelling” — need to feed this customer view and, as analytics will help determine, time triggers and structure messages from this shared customer snapshot. Short of this grand view, at the very least, no part of the organization should conflict with or confuse customer messaging.
A customer-centric merchant must use comprehensive data, and develop all marketing strategies using behavioral, attitudinal, and marketplace data.
Relevant customer messaging also requires listening skills. Marketing is no longer limited to targeted outbound communication. It is a dialog, and brands must be adept at creating listening posts that identify individual customer preferences, problems, and opportunities — more customer data to record, accumulate and analyze. An integrated data platform must track “dialog” data elements, such as important gift giving dates, style preferences, product interests, persistent reasons for product returns, and so on.
Marketers need to view customers as active players — co-creators of value. Consumers today already expect personalization — they assume two-way communication in all situations, and they demand the ability to communicate via their preferred channel. When customers collaborate — and they do — there will exist a dialog across multiple channels, and an abundance of new customer-level data just begging for action-oriented customer insights.
USING DATA TO REDEFINE THE ROLE OF THE RETAILER
The focus of one-to-one marketing beyond the store has often been direct mail, used predominantly as a one-way communication vehicle. Direct mail remains a workhorse in generating store traffic, customer by customer, but it also supports two-way communications, soliciting direct orders via catalogs.
The next generation — data-based marketing — has sought to group like individuals together, a useful exercise necessary for analytics, but still one that has not traditionally incorporated today’s “dialog” data. The limitations of these two approaches are significant: unsophisticated technology, limited customer data, the false notion that “more is better,” and the fact that success is based on the profitability of individual campaigns, rather than “shared value” and profitability per customer.
Going further, all data points have not been used to deliver the most relevant content and offers to each individual customer. As a result, the same “best” customers are being contacted over and over while all the rest are mailed only once — if at all.
Today’s marketers should be focused on multistage CRM — a partnership between retailer and customer in which ongoing two-way dialoging is critical to success. As data-based marketing has evolved, smarter data usage and data collection has led to the construction of multivariate segmentation.
Successful analysis here requires retailers to revisit some of the very same questions they once asked themselves for single-channel analysis. But this time the answers must incorporate a full view of the customer across the brand experience.
What is the best way to communicate? Who are your best customers? What do they buy, where and when? What don’t they buy? What events or offers would appeal to each? What would help each customer become more profitable over time? Are crucial data elements in place to support accurate answers to all these questions? (See “Tracking Multichannel Customer Data,” left, for more.)
A retailer also must ask hard questions: Do our current systems support customer listening, the tracking of dialog data, customer “truth” based on multichannel interaction — and the analytics to support two-way, relevant communication? And in today’s shifting retail economic landscape, merchants should dig deeply into their data to predict the downstream effect that reduced customer spending will have on their shopper profiles.
For example, mid-tier retailers should look into their data to find those customers who appear to be “infrequent shoppers” but who are about to become “high-spend/highly active” shoppers as their reduced discretionary spending causes them to move from high-end luxury retailing downstream. These consumers may be about to become that mid-tier retailer’s highest spend segment, so they should no longer be marketed to as lower-value, infrequent shoppers.
Most retailers today seem to be headed in the right direction. Thanks to sophisticated technology, they’re building real-time (and robust) databases that house more data, including point-of-service data, gross margin information, demographic and geographic data, store traffic counts, and much more. Customer experiences are being enabled across shopping channels. All customer interaction points are being integrated, and there are new avenues of communication, among them social media, cell phones, and PDAs.
GAINING AN EDGE THROUGH KNOWLEDGE
The first step to developing an analytics-driven approach in a multichannel environment is learning more about customers and associated purchase behavior to help drive relevancy and various resulting initiatives.
Conducting a customer segmentation analysis helps in several ways: laying the foundation for customer-focused communications; prospecting for new profitable buyers; identifying cross-channel shopping opportunities; allocating marketing budgets; and improving customer retention, to name but a few. Such an analysis is also useful for deciding where communications should not take place (or take place differently), to conserve resources more precisely.
Beyond customer segmentation, marketers have other useful data analysis tools to perform. Retailers might also consider a profile analysis, which provides an overview of the customer base in terms of “active” vs. “new customers” (with careful consideration paid to the definition of customer timing), customer spend, and number of visits.
A cluster analysis, on the other hand, could provide a statistically driven segmentation schema to differentiate customers. This type of analysis looks at total purchase behavior and appends demographics to describe differences among clusters.
TAKING DATA ANALYSIS TO THE CUSTOMER THROUGH ‘TRUE’ MARKETING INSIGHT
Let’s look at the basic customer metrics of one particular multichannel retailer. An overview sample might reveal, for example, that the merchant’s best customer:
- Spends $275 annually
- Shops via 2.1 channels
- Visits a physical store two times per year, and spends $79 per visit
- Shops in 2.8 departments and spends $55 per department
- 45% of customers are core/repeat shoppers
- 55% of customers are new this year
- 45% of all customers shopping last year have not returned this year (attrition)
In this same example, “active new” customers have the highest spend-per-visit of all groups at $161. “Active not new” customers have consistent spend/visit across both years. “Inactive” customers had high spend/visit last year, but their average visits were the lowest of all groups.
So what does all this mean? For one, because the lower number of visits for “active new” customers reflects the shorter average period during which they shopped this year, the retailer might test “new customer” triggers with bounce-back offers to keep first-time shoppers engaged, and to help discern their preferences early on.
The retailer in this example followed a three-step segmentation analysis approach:
Used statistical techniques to develop data-driven customer segments.
Used transactional data to formulate the segments (a random sample of “active” customers who shopped via any channel within the latest 12 months and who had at least 12 months of history on the database).
Used transactional data and appended demographic data to profile and describe the segments (media usage and preferences, age, gender, marital status, presence of children, income, education, ethnicity, and occupation — both at household and individual levels).
The merchant learned that grouping customers together based solely on overall spend and visits is not the most effective way. Also, customers within the same spend group can have very different preferences (casual wear vs. home vs. women’s designer fashions). And to understand the various “true” customer groupings, the merchant needed to conduct a more robust analysis.
Analysis also revealed that some segments are more loyal than others; however, each segment has opportunity and all are important to the overall business. For example, the largest group (“the decorators”) mostly shopped only once. Why?
The retailer discovered that this group as a whole shops more than others in the months of May, June, and December. Conducting market research would help determine the different types of communications and timing that might yield a better response among unique customers in this segment.
As another example, the retailer determined that a group consisting mostly of mothers shopped an average of two and a-half times during the past 12 months, and their annual spend was more than average. This group, however, may be the mostly likely to be too busy to shop.
The retailer has e-mail names for many of these shoppers, yet has never encouraged them as a distinct group to shop online or to pre-order online for store pick-up. The retailer might consider how this incentive or service would help to increase overall spend and visits.
Customer insight is a critical component of any marketer’s plan. Embarking upon a data-defined and data-driven strategy that listens to customers, incorporates dialog data elements, and asks and answers sometimes very tough questions can help a marketer uncover “customer truths” at both individual and segmented levels. In a customer-focused environment, these insights enable communications that deliver the results that matter most — relevancy, satisfaction, return on investment (ROI), sales, and profit.
Wendy Lynes is senior vice president, retail markets, for Harte-Hanks (www.harte-hanks.com), a San Antonio, TX-based direct marketing services provider.
Tracking Multichannel Customer Data
The customer is king, and should be the primary focus of any savvy retailer. In developing a customer segmentation analysis in today’s multichannel environments, retailers should first seek answers to these important questions:
- Who are my customers?
- What do these customers look like?
- What do these customers buy?
- Why don’t they buy what they don’t buy?
- Who should be considered a “best customer”?
- What are the key purchasing behaviors or behavior combinations that differentiate the various customers?
- Are these customers shopping in one merchandise category or in many categories?
- What keeps them from shopping additional categories?
- How often do customers shop and what do they spend?
- What keeps them from increasing their spend?
- What are the cross-channel shopping patterns?
- How many customers shop online?
- What are the different opportunities by customer segment?
- How do new customers compare to existing customers in terms of profile and behavior?
- Where are the opportunities for growth in customer value?
Retailers need to have the systems, resources and expertise in place to answer these questions across all media channels, to identify unique individual customers in every interaction, and to analyze and leverage these data.