Salespeople enter revenue estimates for opportunities without much precision with big round numbers such as $1 million. This results in inaccurate revenue forecasts.
A database migration incorrectly converts dates from one format to another. This goes unnoticed for 5 years until a new system begins to use the dates to automatically close unused accounts. This results in customer and compliance issues as accounts are incorrectly closed.
Customers are asked for their email address when they sign up but there is no process to validate, maintain or update these addresses. Samples suggest that up to 30% of these addresses are incorrect or no longer used by the customer.
Products are named and structured differently in sales, support, billing and operations systems. This represents an overhead that makes changes and projects more complex and costly as numerous projects develop different mappings between different product representations.
Customer data has no single owner or process for update and maintenance. It is common for sales, support and billing teams to update customer information in one system without a process to update it elsewhere.
A telecom company has no metadata or audit trail that establishes traceability for change to customer data. If someone updates a discount rate for a customer, nobody can tell who made this change or why it was done.
Master product data is corrupted due to physical data corruption on a disk and this corrupt data is integrated into dozens of systems.
Weather sensors at the same station begin to report completely different numbers. There is no process of sensor competition in place so systems use one sensor as the trusted source without knowing which sensor is malfunctioning.
An organization launches a centralized product catalog to standardize product data. The process of updating the catalog requires agreement between business units. This is politically difficult so teams begin to develop their own catalogs using spreadsheets and fail to update the master data.
A data entry clerk enters the wrong closing date for a mortgage due to human error. This results in a customer issue whereby a bank refuses to transfer the money required to close a sale.
A sales system maintains a foreign key for billing records. This key gets updated in the billing system as part of a project but this change is never reflected in the sales system resulting in a broken data structure.
An ecommerce site asks customers for their age. Customers commonly enter "99", "100" and other improbable ages such that the data looks incorrect.
A finance team calculates an account risk score for customers and store this calculation in a database for use by systems. They regularly change this calculation without recalculating for all customers. This results in inconsistencies and a lack of confidence in the number.
A customer database has no single owner such that multiple business units update data such as address without any coordination or control. A marketing team notices that the customer database doesn't reflect their spreadsheet so they update the database to match. This data is highly corrupt and inaccurate but becomes the de-facto master data for the organization.
| Overview: Data Quality Examples | ||
Type | ||
Definition | The correctness and usefulness of data with respect to its purpose. | |
Related Concepts | ||
