CHURN MODEL FOR PILOT PROGRAM USING LOOKER STUDIO

ANUSH BHARATHWAJ Avatar

ANUSH BHARATHWAJ

CHURN MODEL FOR PILOT PROGRAM USING LOOKER STUDIO

Google Looker Dashboard

About this project

CHURN MODEL FOR PILOT PROGRAM USING GOOGLE LOOKER STUDIO

Problem Statement:

As organizations seek to minimize employee turnover and enhance productivity, it is essential to understand the factors that contribute to employee dissatisfaction and churn.
Leveraging data from a pilot program, this study aims to identify key drivers of employee turnover using a churn model developed through advanced analytics.
The model will help determine how job satisfaction, time spent at the company, workload, and salary levels impact the likelihood of employees leaving.
Insights from this analysis will inform targeted interventions and retention strategies to improve overall employee engagement and reduce turnover rates across various departments.

Overview:

The project entails the development of a sophisticated employee churn prediction model utilizing Google Cloud Platform tools.

Initially, we will build a database in BigQuery to store and manage employee data effectively. Following this, we'll establish a connection to the database using Python scripts run in Google Colab, ensuring seamless data interaction.
With the data accessible, we will utilize PyCaret, an open-source machine learning library, to build and train a churn model that predicts the likelihood of employee turnover based on various factors.
Once the model is refined, the results will be exported back to BigQuery.
Finally, we will visualize the insights through a dynamic dashboard created in Looker Studio, enabling easy access and interpretation of data for decision-making purposes.

WireFrame:

undefined For the employee churn prediction dashboard wireframe, the layout will consist of the following components:

Title and Filters: Positioned at the top, allowing users to filter data by department, time period, or other relevant metrics.
Main KPI Text: Display the overall churn percentage prominently at the center or top of the dashboard as a key metric.
Churn Prediction Model Explanation: A concise description of how the model works, placed below the main KPI, possibly with an infographic or a small diagram.
Supporting KPIs: Additional metrics such as average satisfaction level, number of years at the company, and last evaluation score displayed around the main KPI.
Employee Sentiment Visual: A gauge or emoticon-based display answering "Are Employees Happy?"
Churn Drivers (Bar Chart): A bar chart detailing factors causing churn, such as job satisfaction and monthly hours.
Churn Details (Stacked Bar Chart): Visualization showing the number of employees predicted to stay versus leave across different departments, providing insights into where churn is concentrated.

Conclusion:

The project effectively employed advanced analytics to address employee churn, utilizing a dataset of 15,000 employees to train and test a predictive model.
With a calculated churn rate of 7%, and leveraging tools like Python, BigQuery, PyCaret, and Looker Studio, we identified key churn drivers such as job dissatisfaction and excessive work hours.
The dashboard highlighted departments with higher churn rates and provided actionable insights through dynamic visualizations.
This data-driven approach facilitates targeted intervention strategies to improve employee satisfaction and retention. Ultimately, the project underscores the value of integrating machine learning and business intelligence to mitigate turnover and enhance organizational health.

Recommendation:

undefined Consider implementing the following recommendations:

Employee Recognition Program: Incorporate a feature to track and highlight employee achievements. Recognizing and rewarding employees can significantly boost morale and reduce turnover by enhancing job satisfaction.
Professional Development Initiatives: Add functionality to monitor participation in training and development programs. Promoting professional growth encourages employees to remain committed and satisfied with their career progression.
Retention Incentives: Display metrics related to tenure and implement alerts for upcoming retention milestones. Offering bonuses and incentives for long-term commitment can help retain key talent and reduce churn rates.

Findings from the Dashboard:

The dynamic dashboard reveals several critical findings about employee turnover within the organization. The overall churn rate stands at 7%, with an average employee satisfaction level of 50.16%. Employees with longer tenure and higher satisfaction levels show a notably lower propensity to leave, underscoring the importance of job satisfaction in retention.
The dashboard indicates that the most significant churn occurs in the sales and technical departments, with dissatisfaction and excessive work hours being the primary drivers, as depicted by a bar chart.
Analyzing the stacked bar chart, it's evident that the support and IT departments have a higher prediction of employees staying, suggesting better conditions or management practices in these areas. These insights are crucial for targeted retention strategies

Machine Learning Workflow:

undefined

In the machine learning workflow for predicting employee churn, the process starts with training a machine learning model on 70% of the available dataset, which includes data on 15,000 employees.
This training phase involves using historical data to allow the model to learn and identify patterns that indicate churn. Once the model is trained, it is tested and fine-tuned for accuracy on the remaining 30% of the data.
This step ensures the model's predictions are reliable and effective. Finally, the optimized machine learning model is deployed to predict which employees might leave the organization.
This entire process is managed using ML Looker for visualization and BigQuery for handling and processing the data, creating a streamlined workflow from data management to predictive insights.

ANALYSIS QUESTION

1. What is Causing Employees to Leave?

The primary factors causing employees to leave, as identified by the churn prediction model, include job dissatisfaction, excessive work hours, and lack of recognition. The data analysis indicated that these elements contribute significantly to the decision of employees to leave, especially in high-turnover departments like sales and technical support.

2. Are Employees Satisfied?

The dashboard shows an average employee satisfaction level of 50.16%, which suggests a moderate level of contentment among employees. However, this also highlights a significant portion of the workforce that may not be fully satisfied, pointing to areas where improvements could be beneficial.

3. What Departments Have the Most Churn?

According to the dashboard, the departments with the highest churn rates are sales and technical departments. These areas showed higher levels of dissatisfaction and stress due to workload, which aligns with the reasons employees are leaving.

PROJECT QUESTION

1. What Does Success Look Like?

Success in this project would be reflected by a noticeable reduction in the churn rate, improved employee satisfaction scores, and effective implementation of targeted retention strategies. Success would also be demonstrated by a decrease in turnover in high-churn departments and better employee engagement scores across the board.

2. What Does Failure Look Like?

Failure would involve a continued high churn rate or an increase in employee turnover. It would also be evident if the intervention strategies based on the dashboard insights failed to improve or positively impact employee satisfaction and retention, particularly in critical departments.

3. What Trends are Important?

Key trends to monitor include changes in employee satisfaction levels, churn rates by department, and the impact of specific HR interventions on these metrics. Monitoring these trends will help assess the effectiveness of strategies implemented and guide future decisions.

4. What Actions Affect the Trend?

Actions that could positively affect these trends include implementing robust employee recognition programs, offering professional development opportunities, and refining workload management to prevent burnout. Adjusting compensation and benefits to match or exceed industry standards can also influence retention positively.