Besides the general considerations presented in the previous post, the following are some specific GUIDELINES to use the function point in estimates.

Although some authors quote the use of function points directly to derive initial estimates of duration, defects and team size, the most common use is for effort estimation (usually amount of hours).

The process to estimate effort is very simple: Given a productivity (hours per function point) in a given development ENVIRONMENT, simply multiply it by the functional size of software to obtain the desired estimate.

However, the key question is: which productivity should be employed? Many people use market indicators published by various organizations. But many of these people are frustrated with the outcome.

The ANSWER is: there are no magic numbers. The productivity to be employed is specific to each organization and not a market AVERAGE. It must reflect the reality of the development process of the organization in a particular context: development tool, business area or technology platform.

To obtain its own numbers, the organization may use the data from previous projects and recover information such as effort and size in function points. Grouping similar projects, it is possible to obtain a reliable INDICATOR of productivity.

Besides the general considerations presented in the previous post, the following are some specific guidelines to use the function point in estimates.

Although some authors quote the use of function points directly to derive initial estimates of duration, defects and team size, the most common use is for effort estimation (usually amount of hours).

The process to estimate effort is very simple: Given a productivity (hours per function point) in a given development environment, simply multiply it by the functional size of software to obtain the desired estimate.

However, the key question is: which productivity should be employed? Many people use market indicators published by various organizations. But many of these people are frustrated with the outcome.

The answer is: there are no magic numbers. The productivity to be employed is specific to each organization and not a market average. It must reflect the reality of the development process of the organization in a particular context: development tool, business area or technology platform.

To obtain its own numbers, the organization may use the data from previous projects and recover information such as effort and size in function points. Grouping similar projects, it is possible to obtain a reliable indicator of productivity.