Magazine article Personnel

Tapping Internal Talent with a PC

Magazine article Personnel

Tapping Internal Talent with a PC

Article excerpt

Tapping Internal Talent With a PC

Large organizations often find it easier to search the ends of the earth than to look in their own backyards for job candidates. They may rely on headhunters, classified ads, college placement services, career fairs, and other recruitment methods to find job candidates outside the organization, but to hire from within, many employers rely on no more than word of mouth, a job-posting board, and office scuttlebutt. As a result, finding the right job candidate from within a company can be like finding the proverbial needle in a haystack.

Employers that want to hire from within must actively reach out to their employees. Doing so, however, can be difficult when employees number in the thousands. For American International Group, Inc. (AIG), a worldwide insurance and financial services company, a computerized inventory of employee skills and knowledge has proved to be the solution.

Developing the Database

With some 28,000 employees operating in 375 offices in more than 130 countries and jurisdictions, AIG was finding it increasingly difficult to tap internal candidates for job openings. The human resources department needed a better way to find qualified employees for new openings with a minimal loss of time. It therefore decided to develop an internal help-wanted database that could produce qualified job candidates instantly.

Called the "Manpower Planning Inventory," the system was developed on a personal computer in-house. Thanks to today's personal computer technology, this approach was not only the most cost-effective but also the least time-consuming. By using a personal computer, AIG gained better control of system development and maintenance. It also gained a system that was immune to mainframe downtime.

The application code itself was written in a fourth-generation computer programming language. Unlike COBOL, PASCAL and other third-generation languages, fourth-generation languages have a structure and logic that is similar to English.

The simplicity of the fourth-generation language literally saved AIG six months of development time. And although it is quicker and easier to use, the fourth-generation language does not compromise function. The Manpower Planning Inventory is as powerful as any similar system running on a mainframe.

Obtaining and updating data. The major challenge, however, was not developing the system but getting raw data into the system. The most crucial data came directly from employees by means of a questionnaire mailed to their homes. AIG recognized that hiring people to sit at keyboards and input the data from those thousands of questionnaires would be too slow and expensive to be practical. The company therefore decided to use optical scanning to circumvent the input problem. Besides speeding the input process, optical scanning eliminated the human errors and proofing involved in manual input systems.

Designing the questionnaire. Designing a questionnaire that would capture information in a machine-readable format was the next step. For every question, there had to be a listing of specific possible choices. On the answer sheet, employees blackened the spaces corresponding to their choices. Several employees with extensive experience in insurance underwriting helped develop the underwriting categories. However, the most important insights into questionnaire design came from employees who participated in a test run.

One hundred questionnaires were mailed to employees throughout the country. The survey participants were encouraged to make their criticisms blunt. Their feedback proved to be invaluable and probably saved the whole project from becoming a fiasco.

Take the education section, for example. It listed a number of colleges and universities, the majority of which were venerable East Coast institutions. These choices mainly reflected the fact that AIG is headquartered in New York City. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.