JMIR Publications

ABSTRACT

Background:

Employee attrition poses a significant challenge to human resource management (HRM), costing organizations resources, expertise, and continuity, with losses often equating to six to nine months of an employee’s salary (SHRM, 2022). Gamification—the strategic use of game-like elements in non-game contexts (Deterding et al., 2011) offers a promising approach to enhance retention by boosting motivation and engagement. However, its empirical impact remains underexplored. This study introduces the GEAR model (Gamification-Enhanced Attrition Reduction Model), a data-driven framework grounded in Self-Determination Theory (SDT; Ryan & Deci, 2000) and Expectancy Theory (Vroom, 1964), to quantify gamification’s effect on attrition. Using a dataset of 1,470 employees from a multinational corporation, we apply Random Forest and logistic regression with 5-fold cross-validation to test hypotheses linking monthly income, overtime, and tenure to turnover. Results confirm these as significant predictors (p < 0.01), with engagement as a partial mediator, yielding an 85% accurate model. The GEAR model aligns with real-world successes, such as L’Oréal’s Brandstorm (L’Oréal, 2023) and Deloitte’s gamified onboarding (Deloitte, 2022), which reduced turnover by up to 15%. Unlike prior qualitative studies, this research provides a replicable, evidence-based framework, advancing HRM scholarship and offering practitioners actionable strategies—like gamified rewards and mentorship to retain talent. Future research should explore causal links and contextual variations.

Objective:

To develop and introduce the GEAR Model (Gamification-Enhanced Attrition Reduction Model) as a data-driven framework to quantify the impact of gamification on employee attrition. To identify key predictors of employee attrition (monthly income, overtime, and tenure) and assess how gamification can mitigate their effects. To test the mediating role of employee engagement in the relationship between gamification and attrition reduction using statistical modeling techniques (Random Forest, logistic regression, and mediation analysis). To bridge the gap between theory and practice by integrating Self-Determination Theory (SDT) and Expectancy Theory into a replicable, evidence-based gamification strategy for HRM. To provide actionable insights for HR professionals on how gamified interventions (e.g., rewards, challenges, mentorship) can improve employee retention and engagement. To contribute to the gamification and HRM literature by offering a quantitative, predictive framework that extends beyond qualitative studies.

Methods:

Conceptual Framework: The study develops the GEAR Model (Gamification-Enhanced Attrition Reduction Model) based on Self-Determination Theory (SDT) and Expectancy Theory to explain how gamification influences employee attrition through engagement. Data Collection: A dataset of 1,470 employees from a multinational corporation was analyzed. The dataset includes variables such as monthly income, overtime status, tenure, engagement levels, and attrition status. Exploratory Analysis: A Random Forest model (1,000 trees, max depth 100) was used to identify the most influential predictors of attrition, ranking income, overtime, and tenure as key factors. Statistical Modeling: Logistic Regression: Used to test the direct impact of income, overtime, and tenure on attrition (Hypotheses H1 & H2). Mediation Analysis (Baron & Kenny method): Assessed whether engagement mediates the relationship between these predictors and attrition (Hypothesis H3). 5-Fold Cross-Validation: Ensured the model’s predictive accuracy (85%). Gamification Intervention Analysis: The study infers gamification’s impact through engagement data and discusses real-world applications (e.g., Deloitte’s gamified onboarding, L’Oréal’s Brandstorm) to validate findings.

Results:

Descriptive Findings: Median tenure: 7 years Percentage working overtime: 28% Percentage reporting high engagement (>70): 60% Attrition rate: 16%, highest among early-tenure R&D employees Predictor Importance (Random Forest Analysis): Monthly Income was the strongest predictor of attrition (importance score: 0.35) Overtime ranked second (0.30) Tenure ranked third (0.25) Other factors like age and job role had minimal impact (<0.05) Hypothesis Testing (Logistic Regression Results): H1 Supported: Low monthly income (β = -0.45, p < 0.01) and overtime (β = 0.38, p < 0.01) significantly predict attrition. H2 Supported: Early tenure (<3 years) significantly increases attrition risk (β = 0.61, p < 0.001). Odds Ratios: Lower income employees were 36% more likely to leave. Employees working overtime were 46% more likely to leave. Employees in early tenure were 84% more likely to leave. Mediation Analysis (Engagement as a Mediator): Engagement partially mediates the relationship between income and attrition (Sobel z = 2.95, p < 0.01) and tenure and attrition (z = 2.41, p < 0.05). Overtime’s effect on attrition is mostly direct, indicating gamification has limited impact in this area. Accuracy & Model Validation: The model achieved an 85% accuracy rate in predicting attrition. 5-fold cross-validation confirmed the robustness of findings.

Conclusions:

his study provides a data-driven understanding of employee attrition, highlighting the critical role of monthly income, overtime, and tenure in predicting turnover. Our findings confirm that lower salaries, excessive overtime, and short tenure significantly increase the likelihood of employee departure, emphasizing the need for strategic retention policies. Key takeaways include: Compensation Matters: Salary remains the strongest predictor of attrition, suggesting that competitive pay structures can significantly enhance retention. Overtime & Work-Life Balance: Employees working overtime are more likely to leave, highlighting the importance of workload management and well-being initiatives. Early Tenure Risk: New employees (within the first three years) face the highest attrition risk, underscoring the need for structured onboarding, mentorship, and engagement programs. Moreover, engagement partially mediates the relationship between income, tenure, and attrition, indicating that fostering a positive work environment can mitigate turnover risks. However, for overtime, the effect is largely direct, suggesting that improving engagement alone may not be sufficient to counterbalance long working hours. From a practical standpoint, businesses should focus on competitive salaries, workload optimization, and tailored retention strategies for early-career employees. Future research can explore industry-specific variations and the role of non-financial incentives in employee retention. Clinical Trial: Not applicable (observational study based on existing employee data)