How the normal distribution method of company performance appraisal is applied.

If, the enterprise's objective decomposition is reasonable; the appraisal index setting is reasonable; the appraisal and evaluation criteria are reasonable.

Then, the appraisal results should be reasonable.

The results of the performance appraisal do not mean that everything is fine, there are still such and such problems that plague the managers. For example

Questions.

1, What if after the appraisal and evaluation are completed you find that the total bonus amount is high and exceeds the total amount of the planned bonus pool for the performance bonus?

2. Performance results are the main basis for promotion, how can you make performance results a valid basis rather than a formalism?

Part I: What is the normal distribution method

I. Types of normal distribution method

To effectively control financial costs and budgets, and to control the number of bonuses, many companies often use the [normal distribution method], also called [the forced distribution method]. The normal distribution method is divided into the "proportional distribution method" and the "difference distribution method". The so-called proportional distribution method, that is, the performance is divided into different levels, and each level has a corresponding proportion of the number of employees, to ensure that the average score of performance, to ensure that the amount of the bonus controlled within the bonus pool. This is shown in the table below.

The so-called "differential distribution method" requires the staff's performance appraisal score to meet a predetermined margin of difference to ensure the reasonableness of its distribution, which in practice has proved that the "differential distribution method" is easier to operate than the "proportional distribution method In practice, it has been shown that the "differential distribution method" is easier to operate than the "proportional distribution method" and avoids the phenomenon of "bunching" of appraisal scores.

"Is it easy to demotivate employees?"

This is a normal thought for many managers, and if you consider it from the company's perspective, it ensures that the best employees are honored, thus separating the best employees from the relatively inferior ones. But there is a question we must consider clearly before implementation, since the need for normal distribution, then, who should have the "normal distribution right"?

Secondly, who should have the "right to normal distribution"?

To answer this question, we must first briefly analyze the distribution of appraisal rights in the organization. In management, the right to appraisal (also known as the right to reward and punish) is an important authority for managers, and the exercise of appraisal power is an important means for managers to achieve their management objectives. In actual appraisal, to prevent abuse of authority and affect the fairness of the appraisal results, appraisal authority is usually broken down into appraisal execution authority (authority to review appraisal results) and appraisal supervision authority (authority to approve appraisal results). For example, the appraisal system of many organizations stipulates that the appraisee's line manager has the right to carry out the appraisal and the appraisee's cross-level (secondary) manager has the right to supervise the appraisal.

The basic purpose of implementing the normal distribution method is to achieve fairness and reasonableness in appraisals, not in the appraisal process, but through the adjustment of existing appraisal results. Since the "right to normal distribution" is a right to supervise appraisals, the organization should assign this right to the appraisal supervisor, and this supervisory right can directly adjust the appraisal results, so it is not a process supervisory right (the right to review the appraisal results and to ask the appraisee to make adjustments to unreasonable content in the performance appraisal evaluation), but a final supervisory right (the right to adjust the results). Therefore, it is not the right to supervise the process (the right to review the appraisal results and ask the appraisee to make adjustments to unreasonable content in the performance appraisal), but the right to supervise the final results (the right to approve the appraisal results and make adjustments to unreasonable results). Therefore, the "right to normal distribution" should be given to the final appraisal supervisor. By giving the final appraisal supervisor the right to "normal distribution", the organization's appraisal supervision can be put into practice on the one hand, and at the same time, it can avoid the difficulties of some managers in making a reasonable distribution due to the small number of direct reports. Of course, not all "normal distribution rights" holders can exercise their power correctly and fairly, and the adjustment of performance results often reveals the management ability of the holder.

Part 2: Practical application of the normal distribution method

The proportional distribution method is the normal distribution method that many companies are currently using. In the past years of human resources practice, several problems occur with high frequency, and by summarising them, some of the most significant problems require our attention and solution in our work.

First, is it necessary for companies to carry out "normal distribution"?

I once provided HR consulting services for a foreign company in China. The performance appraisal results of this company were set up in seven levels, namely: A/B/C+/C/C/C-/D/E (A - super excellent B - excellent C+ better C normal C - slightly poor D - poor E - poor). In the early stages, I obtained a lot of performance management materials of the company, and after studying them, I found that the A/B and D/E in the performance results were very unusual. The HR director of the company told me: "It's not even a question of less, A is not there at all, B is less to die for, and D/E is the score given to those who cannot pass the probationary period". I asked, "So how do you identify advanced talent?" The HR director sighed, "Isn't this something I'd like to address now? I've always wondered, is there a need for a 'normal distribution' in our business? Is it appropriate for us to have a 'normal distribution'?"

A normal distribution is like a 'tight band' put on the Monkey King by the Goddess of Mercy. If used well, it can significantly improve the performance management skills of managers and motivate employees; if used badly, it can magnify the sense of unfairness and intensify conflicts between superiors and subordinates. The need for a normal distribution of performance management depends on the quantifiability of the work of employees and the maturity of performance management within the organization.　

1. when the degree of quantifiability of an enterprise's employees' work is low and performance management capabilities are also immature, it is necessary to implement a normal distribution within the enterprise to overcome phenomena such as loose and tight, and convergent effects that tend to occur in performance appraisals, and the distance of the performance distribution should be appropriately drawn out to fully reflect differences in performance achievements.

2. With the development and growth of the enterprise, after the performance management capability is gradually improved, the enterprise should gradually dilute this compulsory behavior through the construction of an excellent performance culture and the strict control of personnel inflow, and strive to transform the correct performance concepts and skills into the internal thoughts and behaviors of employees, and eventually abolish the normal distribution, or even dilute the so-called standardized performance management system. It is like the Bodhisattva removing the tight band from the head of the Monkey King after the successful taking of the scriptures. This is a relatively long process, and until the ideal victory is achieved, we still need to adopt measures like the normal distribution to ensure the fairness of performance appraisals and to continuously strengthen the performance management awareness and management skills of managers at all levels of the company, a stage that most companies must go through. Just as the labor law provides for two renewals and the signing of an open-ended employment contract for those who have worked for more than 10 years, many companies do not understand why the state does this and why it has introduced provisions for the termination of employment contracts for incompetence, just to make companies continuously strengthen the level of performance management. These are the means to strengthen the level of performance management and there should be no resistance.

II. Are there any restrictions on the implementation of the normal distribution?

The XYFY Group is a diversified group of companies in the real estate, building materials, hotels, plastic materials, and entertainment industries, but the HR of this company is not very systematic.

The general manager of XYFY Group always asks very pointed questions. When talking about the implementation of the normal distribution of performance appraisals, he asked me two questions: "Why does the normal distribution fail in some companies? Are there any constraints to the implementation of a normal distribution?"

Through years of experience in corporate human resource management, I first suggested to this general manager some common problems that usually exist in companies that fail in applying normal distribution.

1. Inadequate promotion. Failure to make employees understand the true intent of the normal distribution, which employees mistakenly believe is a new form of punishment, leads to resistance.

2. Treat the normal distribution as a panacea. Expect to use it to solve all problems, but instead, it is a one-trick pony due to the lack of supporting measures.

3. Improper timing. Some companies force the normal distribution when the market fluctuates sharply and employees are emotionally unbalanced, leading to a loss of security, cohesion,d loss of a large number of core staff.

4. Rigidity of the normal distribution scheme. Applying the normal distribution to all employees regardless of their job characteristics and nature, and not flexibly adjusting it according to the organiorganizationalture and business conditions of the enterprise, makes short-term fairness turn into long-term unfairness, employees' passion for work gradually decreases, and the normal distribution loses its effect."

What are the "supporting measures" mentioned in point 2? The normal distribution is a mandatory division of appraisal results into three, six, or nine grades to overcome the "pile-up" phenomenon that occurs in performance appraisals. This division is based on the premise of a relatively fair appraisal, without which the normal distribution would only be chaotic. It has been mentioned that the normal distribution must be accompanied by a "key event recording method", suggesting that the latter is an important condition for the successful implementation of the normal distribution. However, years of practice have shown us that the key event method is indeed very effective when applied to the construction and application of the competency model, but it is somewhat difficult to apply it to the normal distribution, because

1. In daily work, it is often difficult to define and fix the behavior of some jobs in the process. The same job, done by different employees, is often handled very differently in the process. Flexible-minded employees are often unconventional, not according to common sense; rigorous people are stable and steady. Some situations depend on flexibility, and sometimes we have to be careful with what we say and do. Therefore, you cannot use key events to measure employees who play differently.

2. It is not easy to have 1-2 key events in a year, what if there are none? No more reviews? Not.

So the supporting facilities should be as follows.

1) A mechanism for setting regular initial targets. It is the correct source of performance results. I once came across a Japanese company whose staff made goal entries at the beginning of each performance scoring period accounting for more than 80% of the staff, such an assessment is completely putting the cart before the horse, how can its normal distribution be fair and effective?

(2) A perfect performance interview mechanism. Performance interviews can objectively help employees to identify problems and thus produce objective results.

(3) Strictly define the concept of "over-achievement". (3) Strictly define the concept of "over-achievement". For example, if an employment contract is signed with an employee, the employee will normally be given full marks for completing and being accurate within two weeks of joining the company, and it will be necessary to separately define how to exceed 100%, rather than the employee thinking or the leader thinking that it is OK.

Thirdly, is it better to have more layers of evaluation or fewer layers?

A few years ago, I worked with a subgroup of a central enterprise on a performance optimization project. During the communication phase of the performance system revision program, the consulting project team mentioned the "normal distribution", but the HR department of Party A thought that the "normal distribution" part was too complicated. In this regard, Party A and Party B held their own views and there was a stalemate.

Party A believes that there are 5 levels of hierarchy: A-10% B-10% C-60% D-10% E-10%

Party B believes that: there are 7 levels of strata: A≦5% B+≦10% B-≦10% C-≧50% C- ≦10% D+≦10% (D-+E)≦5%

When conducting a normal distribution, the division and number of tiers have a great influence on the final distribution result, if the design has few tiers, resulting in a small difference in tiers but a large difference in coefficients. For example, if Zhang rated B and Wang rated C, the result looks like there is only a one-level difference, but the actual difference in performance bonus is 20%. If the design of more levels, results in a large gap between levels, the coefficient difference is small. For example, if Xiao Zhang rated B+ and Xiao Wang rated C+, although the difference between the levels is two levels, the coefficient difference is also 20%; if Xiao Wang rated B, the difference between the levels is one level, the coefficient difference is only 10%.

At the same time, in the application of this proportional distribution method, the key is to focus on the balance of each level, such as the enterprise's performance bonus pool for the current period of the bonus of 30 million, the upper limit of the controllable range can not exceed 10%, which means that the total bonus on the line can not exceed 33 million, imagine, assuming that in the enterprise there are more than 90% of the staff evaluation results are C + or more, then the total bonus can be imagined but must be But if all employees are rated C, the total bonus is not exceeded, but such a result is meaningless.

This premise, if the evaluation level is designed too little, in the normal distribution will find that the proportion of personnel is reconcentrated in the middle of the rating, this is because there is an employee rated B, wilwhich l corresponds to an employee rating D if the evaluation for the B employee position is director, conceivable companies to balance the evaluation results, there may be a nseveralpervisors or commissioners and other lower positions of the staff evaluation results for D or E, the result will inevitably be the same. The result of doing so will inevitably be a "potpotluckr an increase in staff dissatisfaction, many companies to avoid such a situation, director level and above, the performance of staff rarely scores more than 100%, which is tantamount to quenching the thirst for plums. (In practice, manager-level and above (excluding manager-level) personnel will be separately compelled to distribute)

If you think differently, and design more evaluation levels, although you still need to consider the balance, bee of the many levels, the coefficient difference is relatively small, it is easy to achieve the corresponding number of people in each level ratio, to truly achieve the role of reward and punishment, highlighting excellence. And because there are many levels of design, the description of each level will be more detailed in the appraisal, and it is a good opportunity for companies with rough performance evaluaevaluationseat it as a management reform.

Can the normal distribution be applied to all departments?

If there are less than five people in a department, it is very difficult to divide them proportionally. The final result may be that most people's appraisal scores are concentrated in the middle section, or the 'convergence effect' mentioned above is formed. So, how can this problem be solved well?

There are usually two types of situations to deal with.

Situation 1. A multi-departmental merger situation is shown in the following diagram.

From the diagram we can s, ee that the marketing department subordinate departments in market a, market two in doing a normal distribution there is room for adjustment, but the remaining three departments are not so easy to adjust, especially the new product department only four people, it is clear that can not be a completely normal distribution. Since individual departments cannot be easily adjusted, the adjustment needs to be done uniformly within the marketing department, which requires the marketing director to summarise and measure the performance results of each department and then adjust the normal distribution.

Situation 2. A situation where consolidation is not possible, as shown in the diagram below.

As can be seen from the diagram, the four subordinates in the planning department cannot be normally distributed with the other departments, and the director of the planning department reports directly to the general manager, so it is not possible to do the same as in the first case.

At this point, it would be difficult to use the "proportional step-by-step" method. At this point, companies can use the "differential step-by-step method", which is simpler than the proportional step-by-step method in practice, as illustrated in the following examples.

Case Study

1. Enterprise