Price Sensitivity and Shelf-Pricing Analytics for a Madurai Retail Pharmacy Chain

Price Sensitivity and Shelf-Pricing Analytics is a survey-plus-records study of retail medicine pricing at Aarogya Pharma Plus, a six-outlet pharmacy chain in Madurai (name fictionalised; the pricing problem is typical of Indian retail pharmacy). Retail pharmacy margins are thin, the government caps the shelf price of scheduled formulations while non-scheduled categories are open-priced, and government-backed Jan Aushadhi kendras sell generics at far lower prices — yet most customers cannot say what a tablet should cost. The chain currently prices non-scheduled items by copying neighbouring chemists, with no idea which categories are price-sensitive at all.

For my Anna University MBA Project Work (Regulation 2021, Semester 4), I combined three evidence streams: a structured survey of 200 walk-in customers across the six outlets, twelve weeks of anonymised billing records for about 60 SKUs in four therapeutic categories (analgesics, antidiabetics, antihypertensives, multivitamins), and the NPPA ceiling-price list under the Drugs (Prices Control) Order, 2013 as a regulatory benchmark. The survey was designed, piloted and analysed in IBM SPSS Statistics (Cronbach's alpha, chi-square, t-tests, ANOVA, multiple regression), and the record and price-benchmark analysis ran in Excel. The study ends in a shelf-pricing decision framework: which categories can be discounted, where a generic shelf pays for itself, and whether a private-label pilot is worth attempting.

Retail pharmacies in Tamil Nadu sell under a split price regime — scheduled formulations carry NPPA ceiling prices, but vitamins, many analgesics and other open-priced categories leave the margin decision to the retailer. Aarogya Pharma Plus prices these categories by matching neighbouring chemists. The result: discounts are given where customers were never price-sensitive, prices stay high where sensitivity is real, the generic shelf is under-merchandised because customer attitudes are unknown, and DPCO compliance is discovered reactively rather than by design.

There is a need for a measured, category-level view of price sensitivity — built from a customer survey, the chain's own anonymised billing records and the NPPA ceiling benchmark — that tells the owner which categories can bear a sharper price, where a wider generic shelf will retain customers instead of losing them to Jan Aushadhi kendras, and what a defensible private-label experiment would look like.

Abstract

Indian retail pharmacy operates under a split pricing regime: NPPA-notified ceiling prices bind scheduled formulations, while non-scheduled categories are priced by the retailer. This study measures price sensitivity in that open space for a six-outlet pharmacy chain in Madurai, using a 200-respondent customer survey, 12 weeks of anonymised billing records in four therapeutic categories, and the NPPA ceiling list as benchmark. Analysis in SPSS and Excel produces a category-wise pricing and generic-shelf recommendation set for the owner.

Introduction

Medicines are the one shopping basket where the buyer rarely knows the fair price, and the retailer's margin lives precisely in that fog. Aarogya Pharma Plus (fictionalised) runs six outlets across Madurai with a central owner who sets prices in the open-priced categories by matching competitors. With Jan Aushadhi kendras expanding in Tamil Nadu and customers becoming generic-aware, the owner needs evidence, not instinct, to decide where to defend price and where to lead it down.

Existing practice and research gap

• Open-category prices are copied from nearby chemists; no category-level view of sensitivity exists.
• Customer beliefs about generics are unmeasured, so the generic shelf (three racks today) is merchandised by guesswork.
• Compliance with DPCO ceilings is checked only when a supplier circular arrives.
• Published Indian studies link generic acceptance to demographics, but rarely triangulate stated willingness with actual purchase records — stated intent often differs from what the bill shows.
• Gap: a triangulated study (survey + bills + regulator benchmark) small enough for an MBA dissertation but rigorous enough to act on.

Proposed study

• Survey module (n = 200): price recall, branded-vs-generic willingness (5-point Likert items), switching triggers, monthly spend, demographics; systematic sampling of walk-in customers across all six outlets; pilot of 20 responses; reliability checked with Cronbach's alpha.
• Records module: 12 weeks of anonymised bill lines for ~60 SKUs in four therapeutic categories, coded brand/generic, scheduled/non-scheduled, price band.
• Benchmark module: NPPA ceiling prices (DPCO 2013) and published Jan Aushadhi prices mapped onto the same SKUs.
• Analysis: chi-square tests of willingness against age and income; t-tests/ANOVA on spend across groups; multiple regression on monthly spend; Excel price-band and margin analysis; explicit cross-check of stated generic willingness against recorded generic purchases.

Feasibility

• Technical: Google Forms, SPSS student licence and Excel cover every step; the record template is a spreadsheet the chain's billing staff can fill.
• Economic: the study costs only printing and travel within Madurai.
• Operational: the owner has agreed in principle to the data extract; bills are anonymised at extraction, so no patient identity ever enters the dataset.

In scope

• Six outlets of one Madurai pharmacy chain; 200 walk-in customers surveyed; 12 weeks of anonymised bills for ~60 SKUs in four therapeutic categories.
• Price sensitivity, generic acceptance, spend drivers, DPCO-ceiling compliance check, category pricing framework and a private-label pilot design.
• Semi-structured interviews with the owner and two counter staff to interpret findings.

Out of scope

• Clinical outcomes, prescription appropriateness or any patient-identifiable data.
• Wholesale/distribution pricing; e-pharmacy competition modelling.
• Full-chain private-label development — only a pilot design is proposed, not executed.
• Categories outside the four sampled therapeutic groups.

This is a descriptive-cum-analytical study with a triangulated, cross-sectional design: one survey stream, one records stream and one regulatory benchmark, analysed to answer a single pricing decision. The instrument follows Business Research Methods practice — structured questionnaire with a 5-point Likert section, tested on a 20-respondent pilot, reliability screened with Cronbach's alpha (threshold 0.70), then administered via Google Forms with counter-staff assistance for walk-in customers.

Sampling: systematic sampling — every k-th walk-in customer during selected shifts, stratified across outlets and dayparts; non-customers and minors excluded. Ethics: informed consent on the form's first page, no names or phone numbers collected, bill extracts carry no patient identifiers, commercial data reported only in aggregates, and the questionnaire never asks about specific prescriptions.

The study runs three evidence streams into one analysis layer, and the triangulation check is what makes the conclusion stronger than a survey alone.

flowchart TD
  A["Survey: 200 walk-in customers<br/>6 outlets, systematic sampling"] --> D["Cleaned survey dataset<br/>(SPSS, coded variables)"]
  B["12 weeks of anonymised bills<br/>~60 SKUs, 4 categories"] --> E["Coded records dataset<br/>(brand/generic, scheduled/open)"]
  C["NPPA ceiling list (DPCO 2013)<br/>+ Jan Aushadhi reference prices"] --> F["Excel benchmark sheet<br/>shelf vs ceiling vs generic price"]
  D --> G["SPSS analysis: alpha,<br/>chi-square, t-test, ANOVA, regression"]
  E --> G
  E --> H["Excel price-band and<br/>margin analysis"]
  F --> H
  G --> I{"Do stated generic attitudes<br/>match recorded purchases?"}
  H --> I
  I -->|"Yes — trust the attitudes"| J["Category pricing framework<br/>and recommendations"]
  I -->|"Gap — report and adjust"| J

The output of the study is a decision framework, not a statistic — every sampled SKU lands in exactly one branch:

flowchart LR
  S["Sampled SKU"] --> Q{"DPCO scheduled?"}
  Q -->|"Yes"| C["Verify price at/below NPPA ceiling;<br/>monitor quarterly"]
  Q -->|"No — open price"| P{"Price-sensitive category?"}
  P -->|"Yes"| G["Sharpen price; place generic<br/>alternative at eye level"]
  P -->|"No"| M["Hold price; compete on availability<br/>and counter service"]

Two design points deserve emphasis in the viva. First, the triangulation gate: if customers claim generic willingness but bills show branded purchases, the recommendation weights behaviour over stated intent. Second, the compliance branch runs first — no pricing cleverness is permitted on a scheduled SKU; the ceiling is a hard constraint, and the framework treats it that way.

• A discrete-choice experiment (conjoint) on a subsample to estimate willingness-to-pay curves per category instead of segment-level sensitivity.
• A live pilot: implement the pricing framework in one outlet for a quarter and diff-in-diff against a control outlet.
• E-pharmacy benchmarking — extending the price comparison to online platforms to test the shelf-price ceiling the internet imposes.
• A private-label feasibility business case with supplier quotes and break-even volumes for the two best categories.
• Queue and service analytics, since the framework repeatedly shows service, not price, drives the non-sensitive categories.