India's retail sector currently employs 35 million people, roughly 8 percent of the total workforce, and is the country's second-largest job creator. Organized retail alone is projected to reach $230 billion by 2030, and the sector is expected to add 25 million new jobs over the next five years.

However, the sector's growth trajectory sits alongside a persistent retention problem. Attrition rates in organized retail have historically ranged between 22 and 51.5 percent annually, well above the national average of 17.1 percent across industries in 2025. For frontline workers, rates can be nearly three times the cross-industry norm. Replacing a frontline employee costs approximately 40 percent of their annual salary, — and that estimate covers only the direct costs. At the scale of a multi-city retailer, the additional impact on operations, service quality, and institutional knowledge is substantial. The cost, moreover, of losing a productive and efficient employee can be even more devastating, with spillovers into team morale. Retail employees, on the other hand, navigate a cutthroat sales landscape, challenging customer interactions, broken growth ladders 

Yet the default response at most organizations remains the same: recruit faster. However, the instinct to identify the foundation of such high attrition remains highly underexplored in comparison. 

the measurement problem: why attrition is harder to study than it looks

While attrition rates are easy to count, attrition drivers are not. The number tells you how many people left while shedding very little light about why, under what conditions, or what would have had to change for them to stay.

Standard exit surveys have their limits. Exit interviews, when conducted by company HR, tend to be plagued by biased responses. Moreover, most administrative analyses of exit interviews stop at single-word summarizations, with the most significant insight being, "Left for a better opportunity,” with little insight into how or why the person may have been on the lookout for or accepted a better opportunity in the first place. Especially in organizations where feedback channels are opaque or non-existent, exit interviews become more of a compliance norm than a mechanism to institute change. Aggregate satisfaction scores, meanwhile, can mask significant variation by role, store, city, or management quality. However, employee satisfaction remains crucial: the happiest workers can not only improve customer experience but are also better at their jobs. Satisfied employees stay longer, and employees with more experience make fewer errors, developing pattern recognition to solve unexpected problems. A parallel challenge is how to ensure we’re measuring this well enough to uncover the drivers of both satisfaction and dissatisfaction.  

Research on retail attrition adds another layer of complexity. In most industries, one or two factors drive the bulk of departures. In retail, the top six or seven factors each carry significant independent weight: career development, compensation, management quality, workload, recognition, and communication gaps all contribute meaningfully. This multi-factor structure means that fixing one lever in isolation rarely produces durable change.

The implication for research is that you need a method capable of capturing multiple drivers at once, surfacing the interactions between them, and distinguishing structural problems from contextual ones. That is a harder methodological task than it first appears, especially in fast-moving environments such as front-line retail. 

For instance, Goob Business Lab’s work with a leading apparel retailer attests to these arguments. On the surface, many employees reported positive workplace experiences and expressed genuine pride in the brand they represented. They spoke of supportive managers, strong relationships with colleagues, and a sense of belonging at work.

Yet despite these positive experiences, the retailer continued to grapple with persistently high levels of frontline employee turnover, driven largely by voluntary resignations

This was the puzzle the study was built to solve: not just why people leave, but why they leave when, by conventional measures, so much seems to be working. 

building the evidence base: a three-source method

GBL’s Design Team applies a human-centred lens to research, helping firms identify gaps in their systems and co-create recommendations that can be implemented and scaled. In this study, Deborah Charles, Ragaviveka G., Anaya K., and Ushmita S. brought together four concurrent streams of evidence: secondary research, administrative data, qualitative interviews, and a quantitative survey. The team describes this as a “triangulated approach,” with each source contributing a different perspective on the problem.

As Deborah Charles explains, “Data triangulation is essentially bringing in multiple sources of evidence so that we can bring some more objectivity into the process.” While none of the sources on their own provide causal evidence, bringing them together allows researchers to examine the problem through different lenses and capture its nuances.

Firstly, administrative records supplied the baseline context, such as demographic variables, tenure, designation, store location, compensation increments, and exit data. They gave researchers a map of who stays, who leaves, and what the organization already knew about the patterns from more than a thousand current and past employees. 

Then came the survey that added scale. More than 500 responses were collected from stores across five states, with the team measuring employee satisfaction, social support, career growth perceptions, grievance redressal, and workplace pressure.

But numbers compressed the experience. The qualitative layer added depth: 77 one-on-one and group interviews were conducted across 11 stores in four cities, providing the texture that most inquiries lack. Researchers sat with back-of-house staff, frontline sales staff, and store managers, listening for what surveys cannot always surface: the logic behind decisions, the emotional weight of unacknowledged effort, and the gap between a policy on paper and what gets practiced on the floor.

The goal of bringing these streams together was not to produce a ranked list of problems. It was to find the intersections: where qualitative frustrations showed up as measurable patterns, where survey data pointed to something that interview responses could explain, where the numbers and the stories told the same thing in different registers.

behavioral theory on the shop floor:  why use the Theory of Planned Behavior to study attrition?

One of the more distinctive methodological choices in this study was the application of the Theory of Planned Behavior to analyze the qualitative interview data. The framework, developed by psychologist Icek Ajzen, holds that behavior is primarily shaped by the intention to perform the behavior, and that intention is influenced by three factors: an individual's attitude toward the behavior, the subjective norms they perceive around them about the behavior, and their sense of control over whether the behavior is actually possible.

Applied to attrition, this means the decision to leave a job is not reducible to a single cause. It is a product of how an employee feels about leaving, what they believe their peers, family, or industry expect of them, and whether leaving feels like a real and available option given their circumstances.   Deborah - “We were keen on using a behavioral framework to structure our insights by not just an individual’s perceptions about attrition but also the system around them and how that affects their perceptions about leaving.”

The framework’s application to frontline retail attrition in the Indian context is considerably less common. Most attrition analysis operates on a deficit model: it catalogs what is wrong and assumes that fixing the deficits will change behavior. This framework takes a different approach.  Instead of asking only  what is wrong, it asks what is shaping the employees’ intention to leave. Those are related but not identical questions.

In practice, this meant analyzing interview transcripts not just for what employees said about leaving, but how they framed the possibility. Was leaving described as a preference, a necessity, or an inevitability? Did store managers talk about turnover as something within their influence, or as something beyond their control? Did employees who reported satisfaction still express intent to leave, and if so, under what conditions? Such questions shaped both the analysis and the eventual direction of solution design.