Price Elasticity of Demand: A Safe Ecommerce Test Plan
Elasticity is useful only when the test isolates price and protects contribution, inventory, and customer trust.
Opening Answer
A safe ecommerce price test measures how one SKU’s demanded quantity changes when its price changes, while keeping promotion, availability, traffic, assortment, and measurement rules as stable as possible. The result should support a guarded decision, not justify a price change with a neat-looking formula.
Price elasticity of demand becomes unreliable when a promotion starts, inventory runs out, traffic quality shifts, or the product changes position at the same time. A formula computed across promotion, stockout, and traffic changes is not elasticity; it is confounding dressed as precision.
SellerTrove should use elasticity alongside contribution margin, operational limits, customer-facing claims, and predefined stop rules. Start with the formula, calculate the comparison consistently, then evaluate revenue and contribution together.
Table of Contents
- What is the formula?
- How do you calculate a midpoint example?
- What makes an ecommerce test valid?
- How do you interpret revenue and contribution together?
- Which stop rules protect the store?
- Sources
- FAQ
What is the formula?
Price elasticity of demand equals the percentage change in quantity demanded divided by the percentage change in price:
Elasticity = percentage change in quantity demanded / percentage change in price
Quantity demanded is the number of units customers purchase under the relevant conditions. Price is the customer-facing selling price used for the comparison. Because quantity usually falls when price rises, elasticity is often negative.
Demand is commonly called elastic when the absolute value of elasticity exceeds 1, inelastic when it is below 1, and unit elastic when it is near 1. Elastic demand changes proportionally more than price; inelastic demand changes proportionally less.
The sign explains direction, while absolute value supports classification. An elasticity of -2 means a 1% price increase is associated with an approximately 2% decrease in quantity, assuming the comparison is valid and other important conditions are controlled.
Elasticity describes demand response, not business quality. It does not prove that a price is profitable, operationally safe, or appropriate for the customer promise. See this price elasticity reference.
How do you calculate a midpoint example?
Use midpoint percentage changes for a two-point comparison. This method gives the same percentage change regardless of which point is treated as the starting point.
Percentage change in quantity = (Q2 - Q1) / ((Q1 + Q2) / 2)
Percentage change in price = (P2 - P1) / ((P1 + P2) / 2)
Then divide the quantity change by the price change.
Suppose a SKU changes from $20 to $24, while comparable demanded quantity changes from 100 units to 80 units.
Percentage change in quantity = (80 - 100) / ((100 + 80) / 2) = -20 / 90 = -22.2%
Percentage change in price = (24 - 20) / ((20 + 24) / 2) = 4 / 22 = 18.2%
Therefore:
Elasticity = -22.2% / 18.2% = approximately -1.22
The absolute value is about 1.22, so this example is elastic. It illustrates the calculation only; it is not evidence about an actual SKU. See this elasticity calculation reference.
What makes an ecommerce test valid?
A valid test changes price while preserving a credible comparison environment for the SKU. Define the following before launch:
| Element | Guardrail |
|---|---|
| Unit of analysis | One SKU or clearly defined SKU group |
| Price treatment | Documented regular, discount, or test price |
| Demand measure | Units ordered, with cancellations and returns handled consistently |
| Time window | Predeclared start, end, and review cadence |
| Availability | No material stockout or fulfillment interruption |
| Promotion | No untracked coupon, campaign, bundle, or placement change |
| Traffic and mix | Record channel, volume, geography, device, audience, and eligibility |
| Exclusions | Document launches, outages, fraud spikes, and unusual events |
The cleanest comparison changes one principal pricing variable at SKU level while keeping the product, offer, and availability stable. If different customer groups see different prices simultaneously, define treatment assignment before observing outcomes. If a sequential before-and-after test is unavoidable, record concurrent changes and treat the result as less certain.
A stockout can make quantity appear lower even when customers would have purchased more. A promotion can make quantity appear higher even when the displayed price is not the causal driver. A traffic shift can change conversion because visitors differ in intent. These factors can reverse the interpretation.
Retain the price seen, timestamp, SKU availability, order quantity, refunds, promotion state, channel, and relevant traffic context. Definitions and limitations should be repeatable and transparent, consistent with this measurement guidance.
Review customer-facing claims as well. Do not say that a price “increases demand,” “proves willingness to pay,” or “guarantees savings” when the test cannot support that conclusion. Advertising claims should be truthful, non-misleading, and supported by appropriate evidence; see this small-business advertising guidance.
For the broader workflow, connect the analysis with SellerTrove’s pricing category and competitive pricing strategy.
How do you interpret revenue and contribution together?
Revenue response and profit response are different questions. SellerTrove should not approve a price change from revenue alone.
Revenue is:
[ R = P \times Q ]
Contribution is the money remaining after variable costs attributable to the sale:
[ C = (P - V) \times Q ]
Here, (V) may include product cost, payment fees, fulfillment costs, marketplace charges, variable service costs, and other directly affected costs.
A price increase can reduce units while increasing revenue. It can also increase revenue while reducing contribution if fulfillment, returns, discounts, or service effort become more expensive. A price decrease can increase units and revenue while weakening contribution further.
| Outcome | Revenue | Contribution | Interpretation |
|---|---|---|---|
| Units fall less than price rises | May increase | Must be calculated | Potentially attractive, not automatically profitable |
| Units fall more than price rises | Likely decreases | Usually pressured | Strong warning |
| Units rise after a price cut | May increase or decrease | Depends on variable cost | Check break-even volume |
| Traffic or promotion changes | Unclear | Unclear | Do not attribute response to price alone |
Use the discount break-even calculator and contribution margin calculator to structure the economics. Keep elasticity as the demand-response measure, and organize the wider pricing stack through stack-builder.
Which stop rules protect the store?
Set stop rules before the test begins. Practical rules include:
- Stop if the SKU approaches a defined stockout threshold or fulfillment capacity limit.
- Stop if the product, bundle, shipping promise, or checkout experience changes.
- Stop if an unplanned promotion, coupon, campaign, or placement change materially affects exposure.
- Stop if tracking breaks, price display is inconsistent, or treatment assignment cannot be verified.
- Stop if cancellations, refunds, fraud, or customer complaints exceed a predefined threshold.
- Stop if contribution falls below the approved floor, even when revenue rises.
- Stop if a competitor price change alters the reference price or customer choice set.
- Stop if comparable demand is too limited for a responsible decision.
A stop rule does not automatically mean the test failed. It means the original interpretation is no longer safe. Record the trigger, preserve the observations, and label the result interrupted or confounded rather than forcing a conclusion.
Before launch, document the control price, treatment price, eligible traffic, timing, primary demand metric, contribution metric, and decision thresholds. During the test, review guardrails on a fixed schedule. Afterward, calculate midpoint changes, inspect confounders, compare revenue with contribution, and state uncertainty plainly.
Sources
FAQ
What does elasticity of minus 2 mean?
An elasticity of -2 means quantity demanded changes proportionally about twice as much as price, in the opposite direction. A 1% price increase is associated with an approximately 2% quantity decrease under comparable conditions. It is elastic by absolute value.
Why use absolute value?
Absolute value makes classification easier because the usual demand relationship produces a negative number. Values such as -0.5 and -2 indicate inelastic and elastic demand respectively. Keep the negative sign when explaining direction.
How long should a price test run?
Run the test long enough to capture the SKU’s normal purchase cycle and enough comparable demand for a useful decision. Predefine the window rather than choosing a duration to obtain a preferred result, and account for seasonality, promotions, stock, and traffic changes.
Can competitor price changes invalidate the result?
Yes. A competitor price change can alter customer alternatives and shift demand independently of SellerTrove’s price. Record the event, assess whether the comparison remains credible, and stop or qualify the result if the competitive change materially affects the SKU’s choice set.
We track pricing and new tools across the whole catalog. Get an email when prices move or a better tool launches.