Price Calculation

impact simulation

Calculating quickly doesn't mean predicting correctly

Many pricing engines are marketed based on their speed: thousands of prices recalculated in just a few minutes, a formula applied instantly to an entire catalog. This is a real improvement over manual spreadsheet calculations—but it’s also a promise that says nothing about the quality of the resulting decision.

A calculation answers a narrow question: What is the price resulting from this formula when applied to this data? An impact simulation answers a broader and more useful question: What will happen if this price is actually implemented? These two questions are not the same—an engine may excel at the first and remain silent on the second.

What an impact model should include

An impact simulation is not limited to simply recalculating a price differently. It projects the likely consequences of that price across multiple dimensions simultaneously.

Without these four dimensions, a “calculated price” remains an untested hypothesis—no matter how quickly it can be generated. Our article on cross-elasticity, cannibalization, and the halo effect details the statistical mechanics behind the second dimension, which is often the most overlooked.

Price Calculation vs. Impact Simulation

Our article on pricing simulation for testing pricing strategies: It covers the process of testing a pricing strategy as a whole—this guide focuses specifically on what distinguishes, decision by decision, a pricing engine that calculates quickly from one that makes accurate predictions.

The financial implications: a 1% price increase, an 8% profit increase

The importance of accurately forecasting a pricing decision is no trivial matter: structurally speaking, price remains one of the most powerful drivers of the income statement—far more so than volume or costs, given a comparable level of management effort.

This ratio explains why a poorly anticipated pricing decision can destroy value disproportionate to the price difference itself: a half-point pricing error, repeated across thousands of SKUs, has a greater impact on earnings than an equivalent cost reduction achieved through far greater effort.

85% of companies know they can do better

This potential is not lost on the companies themselves—the gap lies instead between awareness of the problem and the resources actually deployed to solve it.

This gap between intuition and tools is precisely reflected in the confusion between calculation and simulation: many companies believe they have solved the problem because they have automated the calculation, without having actually implemented the simulation layer that validates its impact before deployment.

Building a Reliable Impact Simulation: The Four-Step Method

An impact simulation is only as good as the data and methodology used to create it. Here is the process that produces actionable results, rather than an appealing but unreliable figure:

• Compile a clean sales history covering 12 to 24 months at the finest possible level (store or cluster), noting any events that disrupted it: promotions, stockouts, weather-related issues, and competing promotions.
• Isolate the effect of price from other variables by using elasticity measured by store cluster, rather than a national average elasticity that smooths out behaviors that vary greatly from one retail location to another.
• Simulate several pricing scenarios around the target price—not just the final price under consideration—to compare the impact on volume, margin, and revenue before making a decision.
• Validate after the fact: Compare the forecast to the actual results once the price has been applied, and adjust the model if the difference exceeds a reasonable margin of error.

A simplified example: simulating a price increase for an organic dairy product

Let’s look at an illustrative example to clarify the concept. A particular brand of organic yogurt sells for 3.20 euros and moves 1,000 units per week across a cluster of urban stores. A standard pricing calculation would suggest raising the price to 3.40 euros to offset a rise in purchase costs, assuming sales volumes remain stable—but that’s where the reasoning ends.

An impact simulation goes a step further. It uses the elasticity measured for this cluster to estimate the likely loss in volume, checks whether competing SKUs in the same department are capturing a portion of the shifted sales, and quantifies the net effect on the category’s margin—not just on the individual SKU. In this example, a volume loss of around 6 to 8% would still be consistent with a positive overall margin gain, whereas a loss exceeding 12% would push the operation into the red despite the higher price. It is this shift—which is not apparent in a simple price calculation—that a properly constructed simulation allows us to anticipate before the product hits the shelves.

Common Errors in Price Impact Simulation

• Confusing correlation with causation: A decline in sales observed during a price increase may actually be due to a stockout that occurred at the same time, not the price itself.
• Applying a single elasticity metric to the entire catalog, even though it varies significantly by category, store format, and customer profile.
• Ignore the cannibalization effect with similar products, which can increase sales volume for one product while causing its direct competitor to lose an equivalent amount, with no net gain for the retailer.
• Simulate only the scenario deemed most likely, without comparing alternatives, which deprives the decision of a true point of reference.

When Impact Simulation Becomes Truly Essential

Not all pricing decisions require the same level of effort in terms of simulation. An adjustment of a few cents on a low-stakes item may only require a standard price calculation. Conversely, there are four situations in which simulation is virtually essential:

• A broad product catalog, where a single price increase applied across the board to several thousand items can lead to cannibalization effects that are impossible to foresee at a glance.
• A sector facing intense competitive pressure, where poor calibration immediately results in a loss of traffic that is reflected in the following month's figures.
• A margin that is already tight, where the difference between a reasonable scenario and an optimistic one can tip an operation from profitable to unprofitable.
• A high decision-making frequency—for example, for product categories that are repriced weekly—where the lack of systematic simulation eventually leads to the accumulation of minor errors that are difficult to correct after the fact.

Conversely, over-simulating low-stakes decisions consumes team time without providing a commensurate benefit. The challenge, therefore, is not to systematically simulate every price, but to recognize in advance the situations where the gap between calculation and simulation can actually be costly. The most advanced organizations even approach the problem from the opposite direction, starting directly with the desired outcome: this is the principle of goal-based pricing.

Additional Questions

Do you need an in-house data scientist to simulate the impact of a price change?
Not necessarily. A dedicated pricing solution can handle the statistical complexity, as long as the business team understands the underlying assumptions and knows how to question a result that seems inconsistent. The real risk isn’t the lack of in-house data expertise; it’s the use of a “black box” tool whose results no one can explain to management.

Does an impact simulation replace a field test?
No, it precedes it and makes it more targeted. A simulation makes it possible to rule out obviously poor scenarios early on and to focus the more time-consuming field test on the two or three options that are truly still under consideration. The most successful approaches, in fact, combine simulation with partial decision automation, as detailed in our article onagent-based pricing.

See also: Price Modeling.

FAQ

Also in this series

• Pricing simulation: testing your pricing strategy
• Cross-elasticity, cannibalization, halo effect
• How does an AI pricing engine work?
• Promotions: Measure the Incremental Impact, Not the Volume Shifted
• Calculating Sales Margin: The Complete Guide

Sources: McKinsey & Company, *The Power of Pricing*, 2003 · Bain & Company, “Is Pricing Killing Your Profits?”, June 13, 2018 · Booper, internal product data (GENIUS Predict, GENIUS Price)

Further reading

• Why Pricing Decisions Should Never Be Left to a Single Person
• Justifying a Price or Explaining Its Value: Why Your Approach Makes All the Difference
• Price elasticity: definition, calculation, examples
• Pricing simulation: testing your pricing strategy
• Calculating Sales Margin: The Complete Guide to Managing Your Profitability
• Pricing in the Retail Sector: What's Changing (and What Isn't)
• Competitor Pricing Monitoring: The Complete Process, from Data Collection to Decision-Making