One of the biggest challenges in retail operations is and always has been getting to know the customer. Of course, this was an easier task to accomplish before mega-stores, strip malls, the Web and multi-channel contact centers. Back in the day of local markets and small corner stores, customers and inventories could easily be managed on index cards and hand written ledgers.
Today, large and small retailers alike are searching for methods and systems to identify what was bought when, and most importantly by whom. Simply scaling-up older systems, such as the index or ledger system obviously will not work when the business model has changed as drastically as it has in the last fifty years. This can be a huge challenge and collecting customer information still has huge payoffs that offset the challenges.
By knowing and understanding customers and their habits, marketers, management and operational departments can analyze and optimize efforts ranging from pricing models, product stock or shelving configurations to store locations and marketing communications. The crux to opening up this world of application of customer data, of course is to be able to collect that first key – personally identifying information.
There are only two types of customers conducting transactions today – first time or returning customers. To most organizations, there is no differentiation, as they all look the same unless active measures are put in place to identify new from old, much less high value from high maintenance. In order to implement any customer relationship management (CRM) strategy or system, identification is the very first step.
With approximately 116 million households and 276 million people in the US, there are no silver bullets in the battle to identify customers, as there are no single universal identifiers readily available to make all consumers unique across the board. Information that comes close, such as Social Security, Driver’s License, or bank account numbers are not readily provided in the retail environment due to security and privacy concerns among other reasons. A flexible system that can utilize multiple pieces of information to identify individuals is now, and will continue to be the retailer’s best bet to be able to complete an accurate profile of the customer over multiple channels.
The first step in building a customer database is to determine what information will serve as an individually identifying key. This ID will serve as the key to storing and retrieving additional information such as customer and transaction history, analysis and marketing communications.
Types of identifying information that can be collected include name, address, or one of many numeric types, such as; phone, checking or bankcard numbers. Returning customers can be a little easier to identify through usage of account, branded store card, personalized coupons, or loyalty program numbers.
Channels such as catalog contact center, eCommerce and mCommerce may depend partially or entirely on phone number or login and password combinations to identify discrete individuals. Web and contact center channels rate extremely high on the ability to identify individuals, in the 90 to 100% range. While this is the goal, the challenge comes in reflecting their type of success in all channels consistently.
The main barrier to building a multi-channel profile of customers lies in the main purchase channel – the infamous and less prestigious “bricks” of “bricks and clicks” - the retail channel.
While most matching and appending of retail customer data historically has been performed by back end batching and processing, technology is now to a point where it can occur in real, or near real time on the front end of the transaction. Additionally, real time processing opens the door to more relevant messaging, delivery of personalized offers and the ability to update or collect missing customer profile elements such as name, home, work or email addresses, phone number.
Identification and uses
There are two methods for collecting identifying information, explicit and implicit inference. Explicit collection includes data entry or scans of personally identifying information as that listed above. Information such as Zip code can be collected quickly and can be used to provide a wealth of inferred information. While this does not focus on the individual, it is a good first step for those retailers not collecting any information today.
The most efficient means, depending on business model and channel is to request and collect personally identifying information at the point-of-sale. Of course, this process can be affected by the value proposition offered in exchange for consumers’ identities.
Depending on the Point-of-Sale configuration, the explicit collection process can be made seamless to the customer by recording information typically used in the transaction, such as bankcard or checking account number. This serves as a robust additional internal match key, but no longer as a source of external matching services.
Historically, retailers would use a process called Reverse Bank Card Appending (RBCA) to identify customers personally. Credit card numbers would be captured from polled transactions, batched and sent to a third party processing company. Matched records would be sent back to the retailer. Of course, this method of identification was controversial due to the sensitivity of credit information that can be attached to consumers through their tender. Ongoing privacy legislation brought this method under legislative scrutiny, causing the major service providers to discontinue it as a service offering.
This method can be effective for identifying up to 60% of returning customer transactions, once the account is linked to a customer name and address. Cash transactions are the exception to this type of seamless identification since cash does not have banking account numbers; other information is required to match these transactions. Coupled with the relative predominance of cash transactions and incomplete identification match totals, bankcard number matching can be favorable as an incremental identification process, but not as the first line method of identification.
Transparent collection can occur by scanning personalized coupons, frequency or loyalty program cards during the transaction. This method is preferable for identifying existing customers, but will not capture information on new customers. Additionally, members may not elect to identify themselves in small transactions due to the negligible benefit to the program. Membership program enrollment rates can range from the fractions of percentage points to 60 – 80% of the total customer base. Due to the inability to capture new customers and existing customers who elect not to enroll, this identification method is also most effective when combined with other identification methods.
The most obvious, common and potentially intrusive process is to request information verbally for data entry during the transaction. Often, retailers opt for implied information in this case, which can be less invasive to the transaction process.
Of the unique identifying data elements listed above, Address often takes too much time to collect in a busy retail environment and spelling variations can limit proper matching. Name and Zip code combinations can also provide deduplication issues.
Phone number is the most time efficient piece of information to collect. While there is a great proliferation of phone numbers, (most consumers have business, home, mobile and faxes) this ten-digit number can quickly narrow the search to the household level if not the individual. With match rates similar to RBCA, phone number is a viable replacement for the disappearing service. The drawbacks to using phone append are similar to other methods, home phone numbers are usually associated with households, not individuals and are also subject to change.
The most effective way for retailers to capture customer identifiers will continue to be a multi-step process and can be used successfully to unlock a treasure trove of information until a time that universal identifiers become commonplace.
The tools to manage customer information are available to the retailer thanks to developments in technology. POS based systems can be used to draw from and feed data warehouses in real time. Technology can help build robust stores of data as well as expedite the usage of information for analysis and even management of multi channel marketing, such as combination mail, email and retail messaging campaigns.
Let smart marketing make your job easier. Create value propositions to encourage self-identification. A good loyalty or frequency program with tangible soft or hard rewards will have at least a portion of the customer base clamoring to make sure their every transaction is identified. Remember that part of the value of a good program will come from relevance, which can only be derived from customer knowledge – which is created by having complete, accurate customer data.
Routine training should be in place to reinforce policies and ensure that those in the position of power – the cashier know how to use the technology to the fullest in order to provide a strong basis for marketing, management and operational analysis. Often it is the business process that needs to be reengineered to properly capture customer data. Make sure that the proper processes are in place; otherwise the entire operation can be affected adversely.
In order to survive and thrive, savvy retailers will do whatever it takes to know the customer. Building a knowledge base, implementing CRM strategies and communicating with the customer are the keys to developing and maintaining a competitive edge in today’s economy. By understanding your customers, you can differentiate yourself from the competition by simply offering relevance, understanding of the customer’s needs and most of all, a great experience!