An attrition model is a mathematical analysis of organizational attrition, like the number of students who drop out before completing a degree or the rate of employee turnover at a corporation. Such models can be important for understanding attrition and taking steps to address it. They can be conducted by organizations themselves as well as third parties with an auditing interest, like government agencies that may be concerned with the offerings at a college or university. Statisticians usually supervise the development of an attrition model.
This mathematical modeling requires a wide range of data on attrition from an organization. This data can include base rates, as well as information broken down by age, level of experience, reason for leaving or joining an organization, and so forth. The more information available, the more a statistician can do with it. A detailed attrition model can provide useful and meaningful data.
The statistician can prepare a series of breakdowns of attrition, looking at how membership in an organization waxes and wanes over time. This can allow people to pinpoint areas of particular interest, like a spike in decisions to leave a company after hiring a new executive. The attrition model will help analysts understand how people move through an organization and may identify specific patterns that could be important. It can also help companies and organizations assess the success or failure of programs like new student recruitment or more hands-on involvement from supervisors.
In addition to modeling existing attrition patterns, the attrition model can be used in forecasting. Given the available information and known patterns, researchers can show how staffing, client, and student numbers are likely to fluctuate over time. Modeling can also allow an organization to explore what might happen if it made changes to increase retention. For example, a phone company could conduct a survey to see how many departing customers would stay if a policy change like a rate reduction was made. This information could be used to estimate the long term customer retention impacts of such a policy to help the company decide how to move forward.
Some attrition models are prepared for in-house use only. They may contain confidential or proprietary information that a company does not want to release generally to members of the public. Others may be printed in annual reports, disclosures, and other documents. In some cases, there may be a legal mandate to discuss attrition; colleges, for example, have to furnish information on their graduation rate for the benefit of prospective students.