Embedded Files
PROJECT DESCRIPTION
The client is a large MNC with 9 broad verticals across the organisation. The client faces a challenge in identifying the right people for promotion and preparing them in time. Under the current process, the final promotions are only announced after a first round of training and evaluation; and this leads to delay in employees' transition to their new roles. Hence, the company needs help in identifying the best candidates at an earlier juncture, in order to expedite the entire promotion cycle.
After carrying out an exhaustive exploratory data analysis and identifying the factors which predict an employee's probability of promotion, the insights and recommendations are delivered to the company to assist them in selecting the most eligible employees for promotion, and avoid losses of time and training resources.
Python Version: 3.8 Packages: pandas, numpy, sklearn, matplotlib and seaborn.
EXPLORATORY DATA ANALYSIS
INSIGHTS
The employees who were recruited via referrals from existing employees were the least often promoted, with 3% of the total promotions. Conversely, almost 42% of promoted employees were sourced via an external company. Other recruitment channels, such as internet sites, contributed 55% of the employees who were promoted.
At least 95% percent of promoted employees have had no more than two training sessions without promotion, and this applies for employees from all 3 hiring streams. Thus, employees who have failed only one or two training sessions should be preferred; more than two are not potentially promotable.
Employees between ages 26 and 39 are more likely to be promoted than other ages, for both men and women..
Employees with performance scores of 3 and 5 have been those most often promoted. Employees with 1, 2, and even 4 are not promoted as frequently
Regarding the level of education; 67% of promoted employees have a bachelor's degree, 37% have a master's degree, and 2% have less than a high school degree. Consequently, having a degree superior to a bachelor's degree does not represent a determining factor to be promoted.
56% of promoted employees have less than 5 years of experience, 33% have between 5 and 10 years, and the remaining 11% have between 10 and 20 years of experience. Thus, it can be seen that employees with 1 to 10 years of experience are those most promoted. This trends holds true for employees from all 3 hiring streams. [Name=Promotions by Experience
A considerable number of employees have been promoted from 4 current evaluation score groups; from 48 - 52, 58 - 61, 70 - 73, and 80 - 84. This finding implies that an employee with a low average or high average score can be promoted.
High KPIs are correlated with promotion, as at least 69% of promoted employees have high KPIs; however the opposite effect exists for awards, as 88% of promoted employees have not won any prize before.
A model that predicts which employee is elegigle for promorion was developed. to lean more click the button
RECOMMENDATIONS
Having identified the employee characteristics most correlated with promotion, the following recommendations are made to help the firm best select employees to train, and to reduce unsuccessful promotion attempts:
1. Prioritize employees who have been hired via the “other” recruitment stream and via external sourcing channels – as opposed to employee referrals – since at least 97% of the total promoted employees come from those sources.
2. Select employees who have no more than two completed trainings without promotion, since the percentage of promoted employees with more than 2 unsatisfactory completed trainings is less than 3%.
3. Not considering employee age to be a criteria of interest for promotion, as successfully promoted employees are currently drawn from all age brackets.
4. Choosing only employees who have performance scores of 3, 4 and 5 could be useful, since this not only selects for higher-performing employees, but also provides motivation for employees to achieve these performance scores.
5. Although the percentage of promoted workers who only have a bachelor’s degree is much higher than those who have a master’s degree, the two educational levels should be considered equally eligible, since this differential is likely explained by the relatively fewer number of total master’s degrees among the total population.
6. The company should carefully evaluate the criteria for employee awards to ensure that these align with promotion criteria. Currently employees with no awards make up 88% of promoted employees. The purpose of employee awards should be both to motivate and signal good performance in general.
1. Prioritize employees who have been hired via the “other” recruitment stream and via external sourcing channels – as opposed to employee referrals – since at least 97% of the total promoted employees come from those sources.
2. Select employees who have no more than two completed trainings without promotion, since the percentage of promoted employees with more than 2 unsatisfactory completed trainings is less than 3%.
3. Not considering employee age to be a criteria of interest for promotion, as successfully promoted employees are currently drawn from all age brackets.
4. Choosing only employees who have performance scores of 3, 4 and 5 could be useful, since this not only selects for higher-performing employees, but also provides motivation for employees to achieve these performance scores.
5. Although the percentage of promoted workers who only have a bachelor’s degree is much higher than those who have a master’s degree, the two educational levels should be considered equally eligible, since this differential is likely explained by the relatively fewer number of total master’s degrees among the total population.
6. The company should carefully evaluate the criteria for employee awards to ensure that these align with promotion criteria. Currently employees with no awards make up 88% of promoted employees. The purpose of employee awards should be both to motivate and signal good performance in general.
CLICK TO SEE MORE ABOUT:
Page updated
Google Sites
Report abuse