Software metrics has been quite an intensive research area for more than two decades, yet its practical applications have been rather limited.
The 70's being the era for software size metrics, introduced lines of code, and Albrecht's Function Points, to measure the size and functionality of a software product.
The size measurement, useful in its own right, serves also as a component in measuring productivity, and as an estimated parameter to estimation models, which were mainly introduced in the 80's.
In the 90's, the main focus is on software processes, and to improve the software processes is commonly agreed to be the solution for software companies fighting against delivery delays, cost overriding, and quality flaws in the end product.
As measurement is inherent in the concept of improvement, in the stream of software process improvements currently undertaken, the number of practical applications of software metrics can be expected to increase.

In the project EUREX - European Experience Exchange, which is an ESSI (European Systems and Software Initiative) dissemination action, we have analysed the experiences reported by 13 companies.

From practical point of view, there are several questions, e.g. how to identify the right metrics to use, how to introduce a metrics programme, and how to keep it alive. Some of these questions are discussed below -

• Measurement, what is it ?
• What is the thing being measured?
• Why is the triangle - the entity, attribute, and metric - needed?
• What types of use are there for the measurement data?
• Validating metrics and prediction systems?
• How to identify right metrics?
• What is good data?
• How to introduce a metrics programme and to keep it alive?