Customer data – the vicious cycle

During our various projects concerning customer data it’s surprising how often we find our clients management and decisions hamstrung by poor data. Whether this is experienced through sub-optimal campaigns, product development, manual workarounds to processes, data errors, etc, or by the customer who experiences rework, random marketing and mixed SLAs, the effects are hard to hide. 

Now we’re first to admit the importance of the structure of data warehouses or the language of data models, but an overly technical view soon limits thinking, analysis and remediation elsewhere. For most companies, customer data quality issues stem from a range of places, which we've illustrated below in the 'data cycle'.

Often the point of data entry is the weakest link - for a vast amount of data we are reliant on manual entry, perhaps entered at the point of sale, or from a form, by a direct sales person or in a contact centre. Since that person’s job is usually to sell to or service a customer, the ‘administration’ part of the role, i.e. entering the customer details, sales value or query, is all too often prone to error. In data quality terms, the results will be blank fields, shortcuts (address = “XX”), invalid entries (postcode = “1999”), duplicates, out-of-date information and so on. Which ultimately means that the organisation can’t provide good service, run effective marketing campaigns or make effective decisions. Other sources of issues often overlooked or not resolved are in reporting and governance; reporting either not being aligned to real business needs and overloaded with duplicate and inaccurate reports, and governance lacking or weak leading to a slow but inevitable degradation of data standards over time.

The solution – to raise our eyes from single sources of issues and break the cycle at all points. Think about:

• A clear business process for entering customer data, which is understood and followed
• Effective KPIs and measures in place to manage this process and ensure reasonable levels of compliance
• Well designed front-end systems that make it easy for staff to enter everything they should (and prevents or catches errors)
• Checking and cleaning regularly, perhaps daily or real-time
• A structured process for reporting, requirements, alignment to strategy and support
• Clear ownership, governance and control.

Do all that or it’s ‘junk-in, junk-out’, regardless of how well the back-end systems are built.