Customer relationship management is an unfulfilled dream for virtually every marketing organization. In a perfect world, all your marketing, merchandising, creative and service initiatives center on each individual customer and result in offering the best product to each customer in exactly the best possible way and time. Maximize return. Minimize waste. Be responsive to your customer. Market intelligently.  

This holy grail is seemingly out of reach. Companies can, however, make it a more reasonable objective by charting a roadmap that corresponds to the analytical challenges of CRM. CRM is more than a marketing, merchandise or creative business initiative. The proper use of analytics can bring actionable decision intelligence to CRM initiatives. The analytical steps leading to effective and successful CRM are the focus of this article. 

Businesses often ask, “Why focus on analytics, and not operational or other challenges?” I believe CRM initiatives fail mostly due to a lack of analytical support. When analytics lead the way instead of being afterthoughts, the chances of success increase dramatically. Analytics quantify the value, which in turn paves the way for marketing, merchandising, creative, service and other areas to justify their CRM initiatives. Businesses need to focus on customer intelligence before planning the execution. If a company can show huge value with analytics affecting smarter decisions, then it can be sure to figure out how to implement those decisions. I’ve seen the most successful CRM projects begin with successful analysis followed by strong marketing, merchandising, creative, service and operational projects. But it must start with analytics.

Data Collection and Integrity 

Not surprisingly, the first step to successful CRM is the data. Collect everything you can from all internal customer touchpoints. Also, append any external data that augments the internal data and keeps customer contact information current and accurate. In addition, collect and store customer contact information and behavioral data at multiple levels. For business-to-business companies, identify and delineate contacts and businesses. For business-to-consumer companies, identify and delineate individuals and households. The old adage about bad data producing questionable results still applies. Any effort to increase or improve data is generally worthwhile. Case studies (from several national multichannel retailers and smaller direct marketers) reveal that companies receive up to seven-fold return on investment simply from improved delivery of the marketing message.

360-Degree Customer View and Revenue Attribution 

Once the data is of good quality, link all the transactional behavioral data for each customer together with a customer ID. It is imperative to capture the enterprise-wide view across all channels, media and brands. It is equally important to capture the entire life history of customer touchpoints from pre-customer activity through their active and lapsed stages. Every customer touchpoint within an enterprise is potentially important in order to paint the complete picture of each individual customer. Countless valuable insights come from this full view.

Next, assign each customer purchase to its triggering marketing event and attribute the revenue to its marketing source. Installing direct links between purchases and the marketing effort should be done where possible while building a set of inference rules for those purchases without a direct link. There is tremendous value in taking the time to make cause-and-effect connections between marketing and purchase transactions, even if the association is not perfect. The immediate value will come from better circulation decisions for marketing campaigns.

Customer Attributes and Preferences

Once you properly evaluate and track the customer view and revenue attribution, summarize detailed customer transactions into customer attributes. At the minimum, take the time to create a basic set of behavioral customer summaries capturing purchase recency and frequency. As soon as these items are captured, integrate product and channel purchasing history into the purchase recency and frequency derivatives. Eventually, expand the customer attributes to hundreds (possibly thousands) by creatively combining the behavioral metrics to form a solid foundation. This will enable innumerable benefits from improved customer selections.

Next, add more complex customer attributes by deriving customer preferences. Typical examples are purchasing preferences for channel, seasonality, price, product and promotion. Using data mining, these preferences are derived from history for active customers and predicted for the rest. Knowing the preferences facilitates targeting the right message at the right time to increase response rates. Many companies short-change or ignore this information formation step and do not take advantage of their customer data. Creative analytics produces huge value with well-defined customer summaries.

Customer Segmentation

With the customer data now summarized, define a behavior-based customer segmentation schema that fits your business. The best schema has a few, distinct segments that capture the varying performance levels of your customers. A simple schema that captures purchase recency, purchase frequency and channel preference is often a great place to start. The goal is a few segments that differentiate multichannel and multibuyer customers into a few purchase recency levels. Then, monitor the quantity and quality of these customer segments and track the state of your business. 

Secondly, define a lifestyle-based customer segmentation schema. This schema is useful for strategic planning and new customer acquisition. Good segmentation is another solid foundation for both strategic and tactical planning of CRM initiatives. Companies of any size or type benefit greatly from good customer segmentation.

Customer List Reporting and Demand Forecasting

Each month, measure the key quantity and quality metrics of your customer list by the newly defined segments. Track business performance against last month and last year, then measure business performance against your plan. This is ultimately necessary to manage your campaigns more effectively.

For each upcoming marketing campaign, forecast its expected performance if every customer were to receive it. This forecast is the basis for determining the proper proportion of customers qualifying for the promotion. Accurate projections of the overall performance will result in smarter circulation decisions. 

Predictive Modeling for Targeted Marketing Campaigns 

Begin building statistical predictive models for individual marketing campaigns. Build separate models for unique customer groups to maximize the predictive power, and build separate models for unique marketing offers to capture all nuances in customer interests. You can predict customer profitability by blending response, order amount, returns and payment models into a profit model. Next, avoid collapsing customers into score segments, which reduce the power of the models. Predictive models often pay for themselves within the first usage as a result of smarter circulation decisions. 

With customer attributes, customer preferences and predictive models in hand, target the marketing efforts with personalization and versioning, making sure to vary the product and/or promotion to fit the customer. Any size or type of company can achieve higher response and revenue, often as much as 10 to 20 percent, from good modeling and targeted marketing. 

Revenue Cannibalization

By this step, you have mastered the contact-centric best practices (given a marketing campaign, who should receive it) and are ready to move on to the customer-centric challenges (given a customer, how do I market to him or her). 

A good first customer-centric step is estimating the revenue impact between the marketing contacts in your marketing plan. Measure the true incremental demand produced by each individual marketing event in the context of its surrounding marketing events. Revenue cannibalization is the amount of revenue one marketing event takes from another when both events are marketed to the same customer. Revenue cannibalization (or its complement, incremental revenue) is a function of marketing event timing and content. Closer proximity and similar content suggests higher cannibalization (i.e., less incremental revenue). The goal is to create a scheduled marketing contact plan with minimal cannibalization and maximum incremental revenue. With a revenue cannibalization model, optimal contact strategies are possible. 

Customer Subsequent Value 

Another key CRM metric is customer subsequent value, the estimated value of each customer for the next 12 months. With trigger marketing events and the corresponding subsequent value, calculate the near-term, future value of your customers. Experience shows this can be the single most important metric that drives CRM efforts. Every business should measure customer value. 

Advertising Productivity 

CRM is all about maximizing advertising productivity (return on advertising). Derive the relationship between advertising spend and revenue for customer segments. First calculate the return on past expenditures, and then project those historical measurements to predict the future returns on spending. Determine the saturation point of your customers and media by locating when the next advertising dollar spent yields little incremental demand. Learning this relationship will result in refined marketing budgets – spending more in some areas and less in others to generate better ROI.

Customer Transitions 

CRM is, by definition, understanding then treating your customers throughout their business life with you. You should track customer movement between segments each subsequent time period. In addition, understand the patterns of customers flowing through different stages in your business – passing from pre-customer to first-time buyer then to multibuyer and multichannel buyer. With these historic transition patterns you are able to generate more accurate strategic plans.

Contact Optimization 

Contact optimization is one of the milestone best practices for CRM. The goal is to pick the optimal subset of scheduled marketing contacts and versions for each customer. Using the previously mentioned analytic tools, you now can optimize the outbound marketing contact strategy for each individual customer. This optimal contact stream has proven to produce millions of profit dollars annually for several direct marketers who have tested and rolled out a contact optimization solution. Case studies from medium to large-sized direct marketers have shown that 5 to 10 percent of advertising expenditures can be removed with less than 1 to 2 percent loss in revenue. These companies, through executing more productive marketing plans, have attained significant value by reaching this CRm milestone. Media Optimization  Media optimization is another significant CRM milestone. You can now optimize the mix of spending across customer groups, marketing media and future time periods. Simulate future business performance under different advertising budgets, customer transition patterns and advertising productivity by running what-if scenarios and estimating the outcomes of different media budget strategies. Also, you can optimize future budget allocation to meet specified financial targets and satisfy business constraints. This optimal budget allocation when coupled with optimal contact streams provides the CRM answer for your scheduled contacts to your customers. 

Next Best Action 

Finally, CRM must handle the unplanned touchpoints initiated by the customer. For this third significant CRM milestone, you serve up recommendations on the next best action to take with each customer. The next best action can be the product category or class likely to be the customer’s next purchase, the best promotional offer or some other best action. Using data mining and predictive modeling, queue up the next best action for customers visiting your Web site or call center. Cross-selling or up-selling to your customers with the best recommendation can be very productive. 

The last three steps (contact optimization, media optimization and next best action) leverage the preceding ones and complete the CRM marketing analytics roadmap. With these CRM milestones, you are ready to focus on the marketing, merchandising, creative, service and operational initiatives to support a successful CRM program. They enable you to truly optimize your CRM initiatives and your business. There are many necessary yet challenging analysis steps on the road to effective CRM. The roadmap begins with getting corporate data prepared, then generating customer information and making intelligent decisions about customer treatment. Few companies can say they have completed the journey, but the better ones have started and are well on their way. The good news is there is value all along the journey; you do not need to reach the end to reap the benefits. You will reap rewards with each step you take. The journey at times may seem daunting, but the destination is attainable and definitely worthwhile.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access