Grocery chain optimized price and assortment simultaneously

A major grocery chain used Reflect's price simulation and TURF optimization to identify which products drove traffic versus margin, and reallocated shelf space based on actual demand elasticity.

Background One of Sweden's largest grocery chains faced a classic challenge: how do you balance price and assortment when both margin and customer traffic are critical? The chain had historically worked with cost-plus pricing and supplier-driven shelf placement, but saw market shares gradually declining in several categories. ## Challenge The chain's category managers had access to point-of-sale data and supplier statistics, but lacked tools to understand how price changes affected demand at the product level. Assortment decisions were made based on gross margins without considering how products interacted with each other: a private label product with high margin could cannibalize a branded product that drove store traffic. ## Reflect's work We started by building a demand curve per product based on three years of point-of-sale data, calibrated against price changes and campaign periods. In parallel, we conducted TURF analysis (Total Unduplicated Reach and Frequency) to measure which product combinations provided maximum customer coverage. ### Price simulation The simulation model let category managers test price changes and see the effect on volume, revenue and margin before anything was implemented in store. The model accounted for cross-price elasticity between products in the same category. ### Assortment optimization The TURF analysis revealed that 15% of the assortment reached the same customers without contributing unique reach. These products could be replaced or removed without reducing customer coverage. ## Results The project led to a reallocation of shelf space in three test categories. Price adjustments based on elasticity data were implemented for 120 SKUs.

Key takeaways

Demand curves per product based on three years of POS data
TURF analysis revealed 15% assortment overlap without unique reach
Price simulation with cross-price elasticity between products
Implementation in three test categories with 120 price-adjusted SKUs

Why price is not linear

Price does not behave linearly. A 5% price increase rarely produces exactly 5% lower volume. The reaction depends on where you are on the price scale, which category you operate in, and which thresholds exist in the consumer's perception.

Why pricing must be top-down

Start by understanding the full price landscape, not by optimizing individual SKU margins. Bottom-up pricing leads to inconsistent price images and suboptimal portfolios.

Price perception and context

The price of a product is never perceived in isolation. It is perceived in relation to alternatives, to category norms, and to the consumer's expectations. Context determines whether a price feels high or low.

Price barriers and thresholds

Price has thresholds, points where acceptance drops dramatically. A single unit of currency can be the difference between purchase and rejection. Identifying these thresholds is crucial for profitable pricing.

Why willingness to pay is the wrong question

The question should not be "what are you willing to pay?" but "what do you accept paying?". Willingness to pay measures a hypothetical upper limit. Acceptance measures real behavior.

Problems with conjoint for pricing

Conjoint captures trade-offs but misses context, thresholds and lock-in effects. It gives an illusion of precision that can lead to costly mispricing.

Problems with AI pricing without context

AI-driven pricing without understanding customer psychology and market context optimizes blindly. The algorithms find patterns in historical data but lack understanding of why consumers react the way they do.

Monadic pricing model

In a monadic design each respondent is exposed to ONE price, not a price ladder. This eliminates comparison effects and yields realistic acceptance data that mirrors real purchase decisions.

Captive demand and lock-in effects

Much pricing ignores that customers are often not free to switch. Lock-in creates pricing room that does not show up in standard models but is crucial for the right pricing strategy.

Reflect pricing framework

Our framework combines monadic price measurement, context analysis, threshold identification and calibration against transaction data. It produces pricing decisions that hold up in reality.

See related service→

Want to optimize price and assortment with data from your own business?

← Back to How pricing models can reach all the way to the right decision

Grocery chain optimized price and assortment simultaneously

Key takeaways

Related articles