When a major plant and seed retailer launched its first data quality programme, it found that 9% of its catalogues were being sent to deceased customers. It was wasting a significant element of its marketing budget and alienating potential customers from direct marketing targeted using poor quality data.
Increasingly, the internet is the place where business is transacted, where people buy and sell things, and where relationships and networks are formed. All these activities depend on the effective creation, collection, storage and use of data. Data is the foundation of informed decision making; but only when the right data is in the right place at the right time.
"Quality data is accurate, complete, reliable, available to users when needed and fit for purpose."
Data that reflects the real world as it's today. If a marketing database that contains a customer's details (name, address, postcode, telephone etc. ) this would be deemed as quality data if all elements of the customer's details are currently true.
All of the data is present and no data is missing. If the customer record did not have a postcode for example, then this record is incomplete.
Data used in organisations, particularly large ones, will often be duplicated and used for different purposes. Reliable data is data that is consistent across all data sources.
If a customer's details are held in various systems to support; marketing, sales, distribution and accounts for example - they should all be the same unless a customer has requested otherwise. An example of this would be having a different address for deliveries or billing.
Data should be available to people and systems when it's needed. If a customer informs the organisation that their address has changed, the organisation should update their records to reflect this in a timely manner. Otherwise, deliveries might be made to the wrong address and/or invoices sent to the wrong address which is likely to impact on cash flow.
Data should be available to all users who have been granted the appropriate rights for legitimate and approved purposes. If a Maintenance Controller is to dispatch a repair engineer to fix a reported fault, they will need to have full details of the customer and the reported fault before they can dispatch an engineer.
This is the most important criteria of data quality. All the previous criteria are ideal requirements. If data was always 100% accurate, complete, reliable and timely, then the data could be considered to be fully optimised and fit for purpose.
According to Gartner Group, over 50% of customer relationship management (CRM) and data warehouse programmes fail or do not achieve business case benefits due to poor quality data. This failure is costing organisations tens of millions in direct investment, management time and lost opportunities.
We have over 18 years experience of managing and cleansing data - it's not a job for everyone as it requires robust techniques and very importantly a great deal of patience. Once the data has been cleansed (sometimes referred to as normalised), it can then be used as a reliable source of data which will engender confidence in internal systems and management reporting and by introducing the Microsoft BI range of products, you'll be able to look for hidden insights that are likely to reveal new opportunities and possible threats.
Also read about the importance of having a robust 'Data Governance' policy embedded in your business.