Customer management: it’s still about the data

Customer management: it’s still about the data
By Tony Fisher, president and chief executive of DataFlux

Published: November 6 2008 16:19 | Last updated: November 6 2008 16:19

Since the inception of the database, companies have been seeking to use collected intelligence to give them a greater insight into their customers. The results of this insight are stronger customer relationships and the ability to better meet their needs.

We can use the retail sector as an example. Based on loyalty scheme data, we might know a customer purchases beer and regularly shops at a time which suggests he works a 9-5 job. This intelligence is then used to tailor marketing campaigns to him, perhaps offering a discount on a certain beer brand or extending free delivery of the goods to his home. The intelligence gained from similar profiles held in the data warehouse are then used to feed broad-reaching marketing campaigns that encourage all customers in each segment to purchase beer and other products at the supermarket chain. This analytical practice has been operating successfully for a number of years and business intelligence products have been guiding these decisions.

Sitting alongside this analytical intelligence of the customer is the notion of consistent interaction. For this consistency, many organisations today have turned to large CRM systems which operate on the premise of organising a company’s operational data to supply customer facing staff with a complete view of the customer so a personal service can be delivered. This would, it is argued, provide the company with the ability to treat each customer as an individual.

By having a consolidated record of previous purchases, customer complaints and personal details, each and every customer-facing employee can be confident they are fully familiar with all the relevant details for each customer interaction. The employee could even predict what the customer might want in advance to avoid customer churn and dissatisfaction.

However, for these sophisticated tools to work properly there is one common element that absolutely must be robust – the underlying data. Without a level of certainty about the data going into these systems, the conclusions drawn from the basic geo-demographic information, such as names and addresses, to the more involved complex behavioral data produced by loyalty card schemes, are rendered unreliable. This in turn means that the overall investment in the systems and schemes cannot be fully realised.

The rule of thumb is that, like most other business assets, customer data depreciates and requires investment to see a return. The shelf-life of a customer database is approximately two years. This means that of 100,000 customer records contained within a database today, only 50,000 would be accurate enough to feed a CRM system for marketing purposes by 2010. Lifestyle changes, such as marriage, divorce, relocation and job termination, must be taken into account. If a customer has more than one entry in the CRM system or database, it can unknowingly lead to multiple interactions with the customer. The real world manifestation of the duplication problem can be seen when a call centre employee doesn’t have the golden customer record and insight they need to resolve a complaint or to cross-sell a new product line.

Similarly, companies often have multiple applications running in disparate silos. The sales system, finance system, CRM system and data warehouse might not be linked in a meaningful way and could each contain a record for “John Smith”. Only by merging this data together can a truly effective understanding of Mr Smith and his requirements be gained.

The most innovative data-oriented customer management companies often rely on customer data in order to remain efficient, drive sales and ensure compliance. These businesses are constantly investigating new approaches to stay ahead of their competitors. Master data management (MDM) – described by some as a “data utopia” – manages data at the master level by creating a single data repository and ensuring the quality of the data throughout an organisation. Using this master repository, it is then possible to “feed” other business systems in real time and overcome the battle against silos. The development of service-oriented architecture (SOA) as a discipline has made it possible to establish data governance business rules that can be re-used across multiple systems. These rules help assure high quality data is delivered to each system, regardless of where it is located.

However, like many other areas of technology, successful data management relies on more than just having the right vendor and a well-engineered IT system. Maintaining accurate and high quality data demands the involvement of the wider business and non-IT staff, along with a clear set of policies. Only by agreeing on an enterprisewide definition of what constitutes a customer record can a company determine what important data should be stored.

It is this data that will be used to build the intelligence which will enable the company to deliver exceptional service, which in turn fosters loyalty and increased competitive advantage. It’s also worth mentioning that the departments that usually best understand a customer’s needs, pains and demands are sales and marketing, not IT. As such, a collaboration of IT and business functions are required to deliver data that makes a difference.

Data management is a continuous process and requires constant attention. But it doesn’t have to happen overnight. Start today by building data quality rules and checks into the business processes. Then, as systems are merged together, establish a real-time data quality firewall that can validate data at the point of entry. This will ensure its value from the time of collection until its relevance ceases. At the end of the process, the customer data will be clean and useful, allowing meaningful interactions with each customer and satisfying both the IT and business sides of the organisation.
Copyright The Financial Times Limited 2008