Key Highlights: What This Case Study Covers
- Best practices in pricing analytics for highly regulated CPG industries, including scenario simulation and elasticity modeling.
- Implementation of hierarchical Bayesian and Lasso regression techniques to quantify own- and cross-price effects.
- Building a future-ready pricing engine that integrates regulatory constraints, retailer dynamics, and consumer switching behavior.
- Real-world application of Python-based simulators and model-ready datasets for real-time decision-making in volatile markets.
- Strategic approaches to scale pricing engines across regions while embedding attribution logic and automation for continuous refresh cycles.
Client Overview
A UK-based global tobacco company and the world’s leading producer of fine-cut tobacco and rolling papers. Combustibles account for nearly 90% of its revenue, and the company maintains a dominant presence in the UK market.
The Ask
An advanced analytics-driven pricing engine capable of simulating price changes across brands, tiers, and retail segments. The goal was to move to a structured, model-based framework that could quantify price elasticity, enable robust scenario planning, and support strategic pricing decisions.
Challenges
- Regulatory Environment: Stringent excise duties and restrictions in a dark market with no open display of products or promotions.
- Pricing Volatility: Competitor-driven price changes demanded agile, reactive decision-making.
- Model Limitations: Pricing models did not capture consumer behavior, retailer dynamics, or regulatory impact.
- Consumer Sensitivity: High price sensitivity led to downtrading and brand switching as prices outpaced income growth.
- Data Gaps: Irregular collection and refresh cycles reduced confidence in pricing decisions.
Our Solution: Centralized Pricing Engine with Real-Time Simulation
Data Integration & Preparation
Unified diverse internal and external datasets, including excise taxes, regulations, and retail dynamics, into a clean, standardized dataset for elasticity modeling.
Elasticity Model Development
Built advanced elasticity models across key accounts, independents, and total market levels using hierarchical Bayesian and Lasso regression. Captured consumer behavior, retailer advocacy, and cross-price effects for decision-ready outputs.
Simulation Tool Creation
Developed a user-friendly simulator enabling real-time scenario modeling across 100+ brand/region/tier combinations. Factored in competitor moves, promotions, and tax shifts for smarter, week-level decision-making.
Pricing Insights & Deployment
Deployed validated models into a modular pricing engine with attribution logic and a semi-automated refresh pipeline. Delivered continuous insights on elasticity, downtrading, and competitive impact.
Impact Delivered
- Elasticity at Scale: Three tiers of elasticity models deployed across key accounts, independents, and the total market.
- Real-Time Scenario Planning: Simulator enabled weekly scenario planning across 100+ brand, region, and tier combinations.
- Consumer Behavior Insights: Downtrading and brand-switching quantified using historical data, conjoint analysis, and loyalty signals.
- Attribution and Market Drivers: Integrated attribution logic to measure the impact of regulations, competitor pricing, and external events.
- Scalable Architecture: Phase-ready engine designed for expansion into international markets with quarterly model refresh cycles.
