Data quality, for a business of any kind and or size, is very hard to define, measure, monitor, and improve. But it matters because the negative consequences of inaccurate information are endless. Addressing data quality requires an global approach which is practically not possible.  The company's data comes in every format and from all directions. Some systems process huge numbers of records or transactions each day. Some systems are batch, some are real-time. Also, a company interacts with other business entities where it has no control over the data quality.

Data quality magament should be an integral part of the data processing pipeline of all segments of IT activities of an company and not limied only to Analitycal systems but have an active role at the very source of creating data every where where it is possible.

Data Quality in a pragmatical sense can be organized in many different ways and there is no universal model that can be applied to every business.  
The minimum requirement however would be Data Profiling and Data Validation and Correction.
Data Profiling is not always neccessary if the information about the data, meta data, is well known. Data Profiling is mostly applied at the very early stages of a project .
Depending on the nature of the business it could be also one of or per demand exercise.
Data Profiling provides information on new data or frequently changing data coming into the system. The Data Profiling results are used to locate the most likely data quality problems and directions for effective Data Validation and correction Rules.
The Data Validation Rules once defined can be compiled into a Data Validation Specification which is then implemented as a task in ongoing production processing.
Data Profiling results, Data Validation Specification and Data Validation Logs are stored into a database for further use.