Reducing Billing-Counter Queue Time in a Neighbourhood Supermarket: An OJT Time-and-Motion and Layout Study

This is a final-year OJT project report for a UGC B.Voc in Retail Management. Our two-member team trained at GreenCart Family Supermarket (a fictional name), a neighbourhood store of about 5,000 sq ft with four billing counters, a fruits-and-vegetables section and a steady evening rush.

During the first weeks of OJT we worked at the front end — bagging, helping cashiers and answering customers — and noticed that long queues, not stock-outs, were the most common complaint. Customers with two items waited behind full trolleys, loose vegetables had to be carried back to the weighing scale, UPI payments sometimes failed and had to be retried, and a few price tags did not match the POS price, forcing the cashier to call a supervisor.

We therefore designed a small time-and-motion study. Using a stopwatch and a simple data-collection sheet we recorded arrival times, waiting times, service times, item counts, payment mode and the cause of any delay for customers at each counter during peak hours. We analysed the data and the store's POS reports in Excel, applied basic queueing ideas (arrival rate, service rate, utilisation, Little's law) and drew a Pareto chart of delay causes.

The report recommends low-cost interventions — an express counter for five items or fewer, a pre-weighing station in the produce section, UPI QR standees at every counter and a daily price-tag check — together with a before/after measurement plan and a standard operating procedure (SOP) for the billing area.

Abstract

Billing-counter queues are a major source of dissatisfaction in neighbourhood supermarkets, especially during evening peaks. This OJT study measures queue behaviour at a four-counter supermarket using a stopwatch time study and POS reports, identifies the main causes of delay with a Pareto analysis, uses basic queueing concepts to explain why queues grow sharply when counters are nearly fully utilised, and proposes low-cost changes with a before/after measurement plan and an SOP.

Introduction

In organised grocery retail the checkout is the last impression a customer takes home. A store can have good prices and full shelves but still lose regular customers to a nearby kirana or a quick-commerce app if the billing line is slow. Retail management courses teach front-end operations, but students rarely get to measure a real checkout. OJT provides exactly that opportunity.

Existing situation (observed during OJT)

• Four general counters; any customer can join any queue regardless of basket size.
• Loose fruits and vegetables are weighed at the counter only if the section scale is busy, and sometimes items arrive unweighed.
• One UPI QR code is taped near the cashier; customers lean over to scan it, and payment failures are retried at the counter.
• Price labels on shelves are updated manually; mismatches with the POS price need a supervisor.
• Staff open or close counters based on judgement, not on any measured threshold.

Proposed study and improvements

• Measure arrival rate, waiting time and service time per counter at peak hours.
• Classify every delay by cause and rank causes with a Pareto chart.
• Estimate counter utilisation and explain queue growth with M/M/c intuition and Little's law.
• Recommend an express counter, a pre-weighing station, QR standees and a morning price-tag audit.
• Define a before/after measurement plan and an SOP for the billing area.

Feasibility

• Technical: a stopwatch or phone timer, printed sheets and Excel are enough.
• Economic: suggested changes cost little — printed standees, one extra weighing scale placement and staff scheduling.
• Operational: the store manager agreed to a trial week, and data collection does not disturb customers because observers stand beside the counters.

At GreenCart Family Supermarket, customers regularly wait a long time at the billing counters during evening and weekend peaks. Shoppers with only a few items queue behind large baskets, loose produce that has not been weighed interrupts billing, UPI payment failures cause retries, and price-tag mismatches force the cashier to stop and call a supervisor. The store manager hears complaints but has no numbers: how many customers arrive per hour, how long they actually wait, how long each transaction takes, and which causes contribute most to delay.

Without measurement, decisions such as opening an extra counter or moving the weighing scale are made by guesswork. The problem addressed in this OJT project is to measure queue time and service time at the billing counters during peak hours, identify and rank the causes of delay, and propose and plan the evaluation of low-cost interventions that can reduce average waiting time without hiring additional staff.

In scope

• One store with four billing counters; observation during agreed peak windows (weekday evenings and weekend late mornings) across the OJT period.
• Stopwatch measurements of queue-joining time, service start, service end, item count, payment mode and delay cause for sampled customers.
• Excel analysis of POS summary reports provided by the store manager (no customer-level personal data).
• Queueing concepts at an introductory level: arrival rate, service rate, utilisation, Little's law and M/M/c intuition.
• Recommendations, a trial plan, before/after table and SOP.

Out of scope

• Installing self-checkout kiosks or new POS software.
• Detailed staff scheduling optimisation or simulation software.
• Collecting names, phone numbers or any ID details of customers — the study is fully anonymous.

Research design: descriptive, observational field study (time-and-motion study) with a planned before/after comparison.

Sampling: peak windows identified from the POS "bills per hour" report (for example 6–9 pm on weekdays). In each observation session each team member watches two counters and records every customer who joins the queue during the session, giving a census of that window rather than a hand-picked sample.

Data-collection sheet (one row per customer)

Delay codes: PT price-tag mismatch, UP UPI failure/retry, LW loose item not weighed, BC barcode not scanning, CH change not available, VR void or return, OT other.

Derived measures

• Waiting time = service start − time joined; service time = service end − service start.
• Arrival rate λ = customers joining per hour; service rate μ = 60 / mean service time (minutes) per counter.
• Utilisation ρ = λ / (c·μ) for c open counters. As ρ approaches 1, queues grow sharply; this is the key M/M/c insight.
• Little's law: average number in queue = λ × average waiting time, used as a consistency check between our counts and times.

Illustrative example (not our data): if 90 customers arrive per hour and the average service takes 2.5 minutes (μ = 24 per hour per counter), four counters give ρ = 90 / 96 ≈ 0.94 — close to saturation, so a small increase in service time causes a large increase in waiting. Your own λ and μ replace these numbers.

Timeline (6 weeks, team of two)