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.
