FPT Software - Preventing Staff Turnover Through Retention Analysis

khoinxm Partner, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Adv Designer, Registered, Frontrunner 2022 Participant Posts: 1 Partner


Nguyen Khanh Bao
Nguyen Dinh Hung
Nguyen Xuan Minh Khoi

Country: Vietnam

Organization: FPT Software

FPT Corporation is a leading global technology and IT services provider headquartered in Vietnam, with nearly US$1.3 billion in revenue and 30,000 employees in 26 countries. As a pioneer in digital transformation, FPT delivers world-class services in Smart Factory, Digital platforms, RPA, AI, IoT, Enterprise Mobility, Cloud, AR/VR, Business Applications, Application Services, BPO, and so on. The company has served over 700+ customers worldwide, a hundred of which are Fortune Global 500 companies in the industries of Aerospace & Aviation, Automotive, Banking and Finance, Logistics & Transportation, Utilities, and more.

Awards Categories:

  • Data Science for Good
  • Excellence in Research

Business Challenge:

According to the article "Employee Retention and Turnover in Global Software Development: Comparing In-house Offshoring and Offshore Outsourcing," published in 2018 by Bass and his colleague, poor employee retention negatively impacts software development productivity and product quality.

At FPT Software, a global company with more than 20,000 employees, hiring and retaining experienced and qualified developers in the job market can be daunting. Therefore, at FPT, we want to understand the factors affecting the employee retention rate, to understand better the potential impact of staff turnover on the software development market so that we can have appropriate solutions.

There are many reasons based on this article for the increasingly high staff turnover, which can be listed as the following:

  • Employment policies
  • Work-life balance
  • Work-life innovation
  • Product quality
  • Long working hours
  • Adverse impact on health (post-Covid effects)
  • Alignment of offshore work hours with onshore (Difference between timezones)

To prepare for this topic, we have collected data from various sources (surveys (exit survey, LinkedIn, social media, as well as other developer forums such as VOZ and reviewcty) to enrich data over the years. ETL manipulation during the preparation stage can be challenging because there aren't many skilled employees, and many of them have developer backgrounds. This is especially true when correcting spellings and normalizing data.

We must retain a team of nine individuals to handle different responsibilities for this topic, as the data crawling and cleaning process needs to be meticulous and well-documented. Fortunately, we have been introduced to Dataiku, a new approach to resolving our issues.

Business Solution:

  1. Our initial impression of Dataiku is that it has a user-friendly interface, is easy to use and manage datasets, and makes creating dashboards, modeling, and feature engineering for data visualization straightforward. The ETL (extract-transform-loading) has also gotten simpler with visual recipes on Dataiku such as Pivot, Join, Prepare, Sync, and Split. Additionally, our data team feel more relaxed and narrowed their attention to model optimization (hyper-tuning parameters and run-time), recipe and plugin development, and task automation.
  2. Dataiku allowed data scientists and employees with less expertise in data engineering to access, store, manage, and transform data prior to deploying Machine Learning algorithms, which allowed us to save a lot of time.
  3. Since we have to collect data daily, managing data files becomes quite challenging if one of our team members is absent or leaves. We can manage all of the labor involved in this scenario, including data collection and storage, workflow and schema consistency checks, ETL, modeling, automation, and the development of multiple dashboards, with just three to four individuals, thanks to Dataiku.
  4. Previously, we used Python or Power BI to execute exploratory data analysis (EDA), modeling using Python, and so forth. Now we can use code recipes or Lab features from a visual recipe. Thanks to Dataiku, we no longer have to produce them independently. This is practical and easy to use.

Value Generated:

Most of the factors behind the increasingly high staff turnover could be summarized as workplace satisfaction. The more the employer keeps this metric high, the lower the rate of employee turnover.

Since the application's initial release in April 2022, my data team has determined which employees are overworked based on their reactions, feelings, and happiness throughout working days. This helps us save a lot of money, time, and effort when providing training with hazy and ineffective programs and courses and assigning experienced mentors while preventing many experienced employees from quitting for unknown reasons.

In addition, it also lets us identify whether the project manager, team leader or HR representative is misusing their position of authority and oppressing employees so that we may implement the proper management procedures.

Value Brought by Dataiku:

Speed and save on resources. The Dataiku UI allows our data analysts to quickly and efficiently perform data cleaning and transformation with ease. Instead of employing a composite team of data analysts and engineers, we were able to save time and human resources i.e. after two weeks of Dataiku training, the team could immediately get started on creating their vision.

Setup Info
      Help me…