You Estimated the Elasticity. Now What?

September 2026 | Applied Economics | Pricing Strategy | ← Back to Blog

An elasticity number tells you how demand responds to price. It doesn't tell you what price to set. This post walks through what actually has to happen between a demand estimate and a real pricing decision.

An Estimate Is Not a Decision

You estimated the elasticity. Now what? It's tempting to treat this as the hard part, but the number itself doesn't tell you what to do. An elasticity estimate describes a relationship. A pricing decision requires an objective and a set of constraints. It also requires a way of comparing outcomes, none of which the elasticity number supplies on its own.

"An elasticity number is an input, not an answer."

The Objective Comes First

Revenue, profit, growth, retention, customer lifetime value, these are not interchangeable targets, and the same elasticity estimate can point toward different prices depending on which one is actually being optimized. A price increase that raises revenue can lower profit if it pushes away high-margin repeat customers in favor of one-time, price-sensitive buyers. The objective isn't a formality to settle before the real analysis starts. It changes what the real analysis is actually solving for.

Substitution Changes the Picture

An own-price elasticity describes how a product's own demand responds to its own price. It says nothing about where that demand goes when a price changes. A price increase on one product can shift demand toward a substitute or change the demand for related products, and a price change rarely stays contained to the single product it was measured on. Any pricing decision built purely from a single-product elasticity estimate is missing this cross-product structure entirely.

Costs Can Reverse the Recommendation

A price change that looks attractive purely on demand response can be the wrong decision once margin is considered, and the two pieces of information are not equally easy to get right. In practice, cost uncertainty tends to matter more for the final decision than elasticity uncertainty does. Working directly with 110,000 e-commerce transactions, a five-point error in cost of goods sold forfeited 70 to 75 percent of the potential profit gain from an otherwise correct pricing recommendation, and flipped the recommendation entirely in some borderline categories. A 20 percent error in the elasticity estimate itself, by contrast, still preserved profitability across every scenario tested. The practical implication is specific: check the cost side of a pricing decision first, before spending more effort refining the demand model, since costs tend to be the weaker link.

Counterfactual Simulation, Under the Actual Constraints

With an objective, a substitution structure, and real costs in hand, the next step is simulating a set of candidate pricing strategies against that objective. This has to happen under the constraints the business actually operates within: inventory limits, contractual price floors, brand positioning, competitive response. Comparing the simulated outcomes across candidates, rather than committing to the first price that looks reasonable, is what turns an estimate into an actual, defensible recommendation.

Validate What You Can

Where it's feasible, the estimate or the resulting recommendation is worth checking against a real experiment before it's fully trusted. An observational elasticity estimate and a randomized price test are answering related but different questions, so agreement between the two is meaningfully more reassuring than either one alone. What's learned from that validation feeds back into the next round of estimation and analysis. So does the eventual outcome of the pricing decision itself. The process doesn't end at a single recommendation.

The Actual Point

An elasticity number is an input, not an answer. What you do with it, against the right objective, the right costs, the right substitution patterns, and the right constraints, is where the actual pricing decision gets made.