Conventionally negative; operators reason with the absolute value. |PED| > 1 = elastic; |PED| < 1 = inelastic.
Price elasticity of demand is a unit-free sensitivity number that says how much demand bends when price moves. The figure is conventionally negative because demand falls as price rises, and operators almost always quote and reason with the absolute value. The threshold sits at one. When |PED| > 1, demand is elastic and a 10% price increase that loses 12% of units pulls revenue down. When |PED| < 1, demand is inelastic and the same 10% increase loses only 6% of units, so revenue rises.
Many consumer categories sit somewhere in the -1 to -3 range in absolute terms, but the variance within that range is what matters. Commoditized categories — basics, consumables, anything where the next tab over sells a near-identical SKU — skew elastic, because the substitute is one click away. Differentiated categories — skincare with brand pull, specialty food, performance apparel — skew inelastic, because the substitute isn’t really a substitute. A useful test: line up two of your own SKUs against that pairing and ask which side each one sits on.
The operator reality is that most DTC brands have never measured their own elasticity and quietly assume customers are more price-sensitive than they actually are. The result is endemic under-pricing. The usual failure mode sounds like: “we tried raising prices once and traffic dropped, so we put them back.” That is a single observation with no control — not an elasticity read. Pricing engines that operationalize this number, including most dynamic-pricing systems, are only as good as the elasticity estimate feeding them.
There are three practical ways to get a real number. The cheapest is a natural-experiment read from a past price change, which only holds when promo cadence and channel mix didn’t shift in the same window — and usually they did. Cleaner is a geo-holdout or matched-market test: change price in a set of markets, hold others as control, and read the unit delta against the counterfactual. Cleanest is a controlled A/B-style price test on a SKU subset — split traffic or stratify SKUs and run the experiment with the confounders pinned down.
One more reframe before the operator decision. “We lowered price and sales went up” is not, by itself, evidence of high elasticity — promo cadence, channel mix, and seasonality were almost certainly moving too. And the actual question is rarely about units sold. It is about contribution margin per unit at the new price point. Walk it through with round numbers: a SKU selling 100 units at $50 with a 35% contribution margin earns $17.50 per unit, or $1,750 in contribution. Raise price 10% to $55 and lose 8% of units to 92, and contribution per unit becomes $55 − $32.50 variable cost = $22.50 — total $2,070, roughly 18% more contribution dollars on fewer units sold. Elasticity is an input to that calculation, not the answer.