Every product has a different relationship with price. Drop the price of a generic phone case by 10% and sales might double. Drop the price of insulin by 10% and sales barely move, because the people who need it were already buying it. The first product is highly price-elastic. The second is not.
Price elasticity = % change in quantity demanded ÷ % change in price
If a 10% price drop causes a 25% increase in units sold, elasticity is -2.5. The negative sign is convention (price up usually means demand down); the magnitude is what matters.
Elasticity tells you which products to discount and which to leave alone. Discounting an inelastic product just gives margin away without driving meaningful extra volume. Discounting a highly elastic product can grow revenue even as the unit price drops.
Most stores treat all products the same way, applying flat discounts across the catalogue. The merchants who win run different strategies on different SKUs, because the data tells them where the price-volume relationship actually pays back.
Run real price tests. Move the price up or down on a single SKU for a defined window, hold everything else steady, and measure the change in units sold. The data is messier than the formula suggests (seasonality, competitor moves, traffic shifts) but a few clean tests will tell you whether a product is broadly elastic or broadly inelastic, which is what you actually need to make decisions.
Example: A homewares store runs a 7-day test. On product A, a 12% price drop causes a 38% lift in units sold (highly elastic, discount works). On product B, the same price drop causes a 4% lift (inelastic, the discount lost money). The store stops discounting product B in promotions and reallocates the budget to elastic SKUs.