Divide employees into three, six, nine

Charlie is worse than Sam, but not as good as Mary. In a nutshell, this is the operational impetus behind mandatory rankings—supervisors categorize employees based on past performance and leadership potential.

In companies that don't rank employees, almost every employee will feel good after a performance review discussion with their supervisor and consider themselves above average, especially when a soft-hearted manager sets the bar too low . In a forced ranking system, however, managers rank subordinates' performance along a bell curve.

Mention this program and people think of General Electric. The company divides its employees into three categories: the best 20%, who receive bonuses, promotions and stock options; 70% "top performers" with good prospects; and the worst 10%. "Companies that place their future on their employees must eliminate the worst 10 percent and continue to do so every year—continuously raising performance standards and improving the quality of the company's leadership." CEO Jack Welch in a message to shareholders in the last letter.

GE is no exception: companies such as Microsoft, Cisco, Hewlett-Packard and Sun Microsystems also use employee rankings. Sun's system is remarkably similar to GE's: 20% "excellent," 70% meet "Sun's standards," and 10% are "poor." HP uses a scale of 1-5, with 15% being the top 5 and 5% being the 1st. Microsoft uses a scale of 2.5-5.

The greatest value of forced ranking is not only in categorizing employees, but also in evaluating the actions taken on each employee after the meeting. The rated best performers (A? level employees) can get rapid development and training, and can be promoted quickly. This ensures that the company is and will be led by a group of top-performing managers. Employees listed in the worst category (C-level employees) generally have to leave their original position or organization.

A common practice in implementing mandatory rankings is for the company to set a small number of criteria for the employees to be evaluated—usually mid-level and senior leadership. At the same time, determine the rating distribution: for example, 20% best, 70% very necessary, and 10% worst. Senior managers then discuss, evaluate each individual against the criteria, and divide them into pre-defined categories.

Five Levels of Analysis Before Ranking

Let's take a look at a company that has completed the mandatory ranking, one of the most well-known consumer packaging companies in the United States and whose parent company is in the top 10 of the Fortune 500 bit. From the beginning, the company has always been moving in the right direction.

The parent company's search for a turnaround by appointing a new CEO found problems: The company was highly profitable, but its market share had not risen for years. The company has a culture of "hire for the long term and do your best," with regular performance reviews, but often too lenient. The succession planning process includes all the recognized standard characteristics, but the same people are rated as high potential for promotion each year—and the organization's key positions are rarely filled by those identified by the succession planning checklist.

The company decides to initiate mandatory ranking, and the process begins, first of all, the company executives understand the role of mandatory ranking in many other organizations. They are focused on finding the components of a mandatory ranking system that, if they implement it, will make the mandatory ranking process fit their culture and goals.

Senior leadership analyzed five key components:

Evaluation Criteria. To assess leadership skills and future potential, standard metrics that apply to employees in all organizational units need to be developed. They developed 4 measurable criteria: Execute with excellence, pursue results with passion, be admired, and make tough decisions.

Organization level. How deep should the ranking process go within the organization? Where will the benefits diminish? Is the manager team also included in the mandatory ranking? To ensure the maximum impact of the program, the company decided to rank the program to include only the top managers and vice president (VP) teams (47 people) and their direct reports (180 people).

Confidentiality. How open is the company's adoption of a mandatory ranking system? Are the assessment results communicated to the assessee? The company finally decided to make the ranking process public to all employees. The principle is "no secrets, no surprises".

steps. How is the assessment done? How long does it last? What is the role of each participant? How should assessors be prepared to understand the business and conduct assessments correctly?

Results and consequences. What happens after the ranking process is completed? How to treat A? level employees? The trickier thing is: what to do with C-level employees?

Main aspects of the ranking process

The process is fully open to everyone directly affected, such as raters and rateees. The company assures everyone that it will disclose not only the final assessment results, but also the strengths, weaknesses and development needs of each individual discussed in the senior leadership discussions.

Each rater is given a booklet with a brief profile of the rateee, including full employment history and statistics from the last two performance reviews. Each attends 3 hours of training to understand the spirit, intent and technical details of the procedure, assessment techniques.

Two weeks before the meeting, the CEO sends two emails to everyone involved in the process. In the first email, he said: "The purpose of the program is to identify the best 20%, the middle 70% and the worst 10% from our leaders. The best 20% career development should be accelerated; The solidarity contribution of the middle 70% is important to our success; the worst 10% of talents are not fully utilized here and may be better utilized elsewhere."

In a second email, he issued a blunt request to all raters: "The future of the company's leadership depends on the 20% you have identified as the best. When identifying, 'exercise managerial talent'. ', careful, careful, selective. We want to identify the best people and create opportunities for them to develop. We also identify the 10% of underperforming people. I'm ready to work with you to ensure that there are Complete the classification work of all employees in order." The evaluation meeting took more than half a week, and each meeting was up to 8 hours. The technical details of the meetings were simple: For each meeting, write the name of each manager to be evaluated on a small post-it note, alphabetically, on a blank flipchart on the wall. A blank graph paper, 5 feet high and 10 feet long, was placed in the center of the room, facing the evaluation group. The paper is divided into three parts with a straight line, the top is written: "best 20%", "very necessary 70%", "worst 10%", and use sticky notes to indicate the exact number of people who should be divided into categories .

Evaluators are trained to get a quick overview of the technical details, rationale, basic guidelines, and key points of the meeting to understand the purpose of the meeting: "Your job is to fully discuss everyone and then move names from the alphabet to A , B, C area."

Uncompromising adherence to principles

A total of 39 people to be assessed. When the raters sat down, they found that the number of people in each area was not close to the desired number. 13 were in block A, 26 were in block B, and block C was empty. During the meeting, the evaluators were first reminded that the company needed results: 7 A's, 28 B's, 4 C's.

The discussion then focused on the 4 criteria previously identified, namely, excellent execution, passion for results, admiration, and tough decisions. They must judge accordingly.

Reducing the list of A?-level employees is difficult because all raters agree that these managers are the best employees in the organization. However, the rules cannot be changed.

Likewise, it is very difficult to place an employee in the C category, even if there is a consensus that an employee is inferior to the rest of the team. The head of information technology said, "I don't want to evict people who don't have replacements, does that mean we have to stop?"

The CEO asked the key question, "If we don't stop, this Is this activity worthless?" Although in the ensuing discussion, departments that have quickly driven out marginal employees are at a disadvantage compared to departments that tolerate mediocrity, but the procedure must be done. "A C-level employee is a C-level employee wherever they are," someone said. The end result is that almost all of the C-level employees on the list come from a department that is consistently tolerant.

The employees who are considered the most likely to be promoted have the longest discussions, but it is difficult to form a consensus.

The CEO said of an employee who has been viewed as the perfect replacement for a manager position for years: "I'm on principle. She's not an A?, she's not forward-looking. Maybe she's A competent manager, but not a leader. Not that she will never be, but it takes a long, hard struggle."

Discuss not only identifying the best people, the people who need it, and people who must leave, but also getting senior management Tiers see development opportunities beyond training seminars and managers' university courses. "Could there be some work at headquarters that would give development opportunities to employees we didn't notice?" asked the vice president of human resources.

The head of operations replies, "There are already several jobs that can be arranged for an 18-month rotation."

During the discussion, the participants talked about temporary work arrangements and special projects for high-performing employees to participate in. Projects, put to the test, get development, and make senior leaders take them more seriously.

Finally, the ranking ends. On the paper there are 7 names in Zone A, 28 in Zone B, and 4 in Zone C. The immediate result is the successful completion of the task: each group of evaluators assigns an appropriate proportion of the population to each category. The more important takeaway was that they had a genuine consensus on the leadership potential of every member of the company's 227 senior managers.

After the meeting, many managers leave exhausted saying, for the first time, they have really learned the depths of the company's best talent pool, recognizing where the peaks and valleys are, and that's perhaps the biggest added benefit of the program.

The original text is reproduced with permission from Across the Board in the November/December 2002 issue of Conference Board Magazine, which is copyrighted by Conference Board. Translated by Huang Xi.

Dick Grote is president of Grote Consulting in Dallas, Texas, author of The Performance Appraisal Question and Answer Book, and founder of GroteApproach, an online performance measurement system.