Glossary/Price elasticity
What is price elasticity?
Price elasticity of demand measures the percentage change in units sold that results from a one percent change in price. It is normally negative — raising price reduces demand — and its magnitude decides everything about pricing strategy. Demand with elasticity between zero and minus one is inelastic: a price rise increases revenue, because volume falls proportionally less than price rose. Below minus one it is elastic, and a price cut can increase revenue. For a marketplace seller the profit version matters more than the revenue version, because a discount has to recover its lost margin from additional volume, and the volume required is usually far higher than intuition suggests.
01/Formula
Formula
Elasticity ε = % change in quantity / % change in price Volume needed to break even on a price cut: required uplift = d / (contribution rate − d) where d = discount as a share of price Profit-maximising markup (constant elasticity): price = COGS × ε / (ε + 1)
Example
A ฿1,000 SKU with a 25% contribution rate, discounted 10% to ฿900. Contribution falls from ฿250 to ฿150 per unit. Required uplift = 0.10 / (0.25 − 0.10) = 0.667 The discount needs 67% more units just to hold contribution flat. That implies an elasticity around −6.7, far more elastic than most marketplace SKUs actually are.
02/In detail
How much extra volume does a discount need?
More than almost anyone estimates, and the thinner the margin the worse it gets. The required uplift is the discount divided by the contribution rate minus the discount, which grows explosively as the discount approaches the contribution rate. A 10% cut on a 40% contribution product needs 33% more units; the same cut on a 15% contribution product needs 200% more; on a 12% product it is mathematically impossible, because the discount exceeds the margin available. This asymmetry is why blanket percentage-off campaigns across a mixed catalogue reliably lose money — the same discount is comfortable on the fat products and ruinous on the thin ones.
required volume uplift to hold contribution flat
discount at 40% contrib at 25% at 15%
5% 14% 25% 50%
10% 33% 67% 200%
15% 60% 150% n/a
20% 100% 400% n/aWhy is elasticity hard to measure on a marketplace?
Because price almost never moves on its own. Sellers cut prices during campaign windows, when the platform is also driving traffic, when competitors are also discounting, and often while the product is being advertised more heavily. Every one of those confounds the price change with a demand change, and the naive estimate — comparing units before and after — attributes the campaign’s traffic to the discount. Marketplace listings also have position effects: a price change alters search ranking and badge eligibility, so demand moves partly because visibility moved, not because shoppers reacted to the number. Credible estimates need price variation that is not aligned with everything else, which usually means deliberate, staggered changes rather than historical data alone.
Are all products on a marketplace price-elastic?
No, and assuming so is expensive. Elasticity varies with how comparable the product is: an identical commodity listed by forty sellers is highly elastic because switching costs nothing, while a differentiated item, a bundle, or a product with strong reviews and few close substitutes can be surprisingly inelastic. Elasticity also varies by shopper segment and by moment — the same SKU behaves differently during a mega-campaign than on an ordinary Tuesday. The practical implication is that elasticity is a per-SKU, per-context estimate with uncertainty attached, and a pricing decision should account for the uncertainty rather than treating a point estimate as fact.
03/Why it matters
The trap, in one paragraph.
Discounting is the fastest lever a marketplace seller has and the one most often pulled without arithmetic. Because platforms surface promotions prominently and reward them with visibility, a discount almost always produces more units, which feels like confirmation. Whether it produced more contribution is a different question, and answering it needs the elasticity and the contribution rate together — either alone is misleading.
Common mistake
Reading a sales increase after a discount as proof the discount worked. Units rising is the expected outcome; the test is whether they rose past the break-even uplift for that discount and that contribution rate. On a thin-margin SKU a 40% unit increase after a 10% cut is a substantial loss dressed as a successful promotion.
04/In DataGlass
How Price elasticity is used in DataGlass.
DataGlass estimates demand response to price per product from order history and surfaces the break-even volume uplift for any proposed discount alongside the estimate, with its uncertainty. Promotion and price recommendations are evaluated on projected contribution rather than on projected units.
05/Sources
- [1] Price elasticity of demand — overview
The standard definition, the elastic and inelastic ranges, and the relationship between elasticity and the profit-maximising markup.
- [2] McKinsey — Dynamic pricing
Retail pricing practice, including the disproportionate profit sensitivity of small price changes relative to volume changes.