What is price elasticity of demand
Price elasticity of demand measures how much the quantity sold of an item reacts to a change in its price: it is the percentage change in units divided by the percentage change in price.
The result is almost always negative, because price and quantity move in opposite directions, and what matters is its size. Above 1 in absolute value, demand is elastic: the drop in units outweighs the rise in price and the item's revenue shrinks. Below 1, it is inelastic: the price can rise without volume reacting in the same proportion.
In retail, elasticity is not a property of the product. It changes by channel, by region, by day of the week, by the presence of a competitor on the same street and, above all, by how well the customer knows that item's price by heart.
How elasticity is calculated
Two percentage changes and one division. The work is to isolate the price change from everything that happened along with it.
percentage change = (new value − old value) ÷ old value
elasticity = percentage change in quantity ÷ percentage change in price
item contribution = (price − cost and variable expenses) × quantity
Before deciding whether the item can take the increase, it is worth knowing how much of it is already yours: the selling price calculator breaks the current price down into cost, tax, commission and margin.
An item that went from R$ 10.00 to R$ 11.00
The item costs R$ 5.10 on the invoice, used to sell a thousand units a month and carries 14% of the price in tax, card fees and commission. The price rose 10% and the volume fell 12%.
| Measured elasticity−12% in volume ÷ +10% in price | − 1.2 |
|---|---|
| Revenue before1,000 units at R$ 10.00 | R$ 10,000 |
| Revenue after880 units at R$ 11.00, or 3.2% less | R$ 9,680 |
| Contribution beforeR$ 3.50 per unit across a thousand units | R$ 3,500 |
| Contribution afterR$ 4.36 per unit across 880 units | R$ 3,836.80 |
Demand is elastic, revenue fell 3.2% and the item's result rose 9.6%. It is not an exception: on a thin-margin item, each unit that no longer sells takes its cost with it, and of the extra real of price only tax and commission take a share — the rest goes into contribution. That is why a pricing decision is not made by looking at revenue: the number that answers is contribution, and it can point the opposite way.
How to measure elasticity without a laboratory
The textbook number is no use: the elasticity of your item, in your channel, is in your sales history.
The measurement is born from an isolated change. Change an item's price, keep display, channel and period comparable, and read the volume for two or three weeks. What invalidates the reading is whatever happened alongside: a competitor's promotion, a holiday, a stockout, a shelf change. Without that care, the measured number describes the holiday and not the price.
Not every item deserves the measurement. A small part of the catalog accounts for the price perception of the whole store — they are the known-price items, the ones whose price the customer knows by heart and uses to judge whether the store is expensive. On them, elasticity is high and the mistake is costly. In the rest of the catalog, the reaction is modest and the room to adjust price is larger than intuition suggests.
There is also cross elasticity: how much a change in one item's price moves the volume of another. It is what describes cannibalization inside the mix itself — raising the price of the leading brand shifts volume to the private label, and the effect on the result depends on the contribution index of the two. Ignoring this relationship is what makes a successful increase on one item show up as a margin drop in the category.
Elasticity is also not constant along the price range. It changes near the steps the customer sees — the jump from R$ 9.90 to R$ 10.50, a category's psychological limit, the point at which the competitor becomes visibly cheaper. Two measurements taken at different points of the same item's range return different numbers, and both are right.
Where reading elasticity misleads
The four below make an operation give up increases that were worth it, or insist on discounts that do not pay.
Deciding price by the effect on revenue
Revenue does not pay the bills: contribution does. An increase that pulls down the item's revenue can increase what is left over, and a discount that makes revenue grow usually reduces contribution — because the extra volume comes in with a lower margin and carries cost. The criterion is always price minus cost and variable expenses, multiplied by the volume that actually went out.
Measuring in a contaminated period
Holidays, a competitor's promotion, stockouts, supplier campaigns and changes in display alter the volume without the price having anything to do with it. An elasticity measured in those weeks is an invented number dressed up as data, and it tends to become a rule for the entire catalog. Without a clean period, it is more honest not to measure.
Extending one item's elasticity to the category
Inside the same category live the known-price item, which reacts to cents, and the tail reference, which barely reacts. Applying the elasticity of the first to the second freezes the price of half the catalog out of fear of a reaction that does not exist — and it is precisely in that half that the margin room the operation is looking for lives.
Treating the number as permanent
Elasticity changes with the price level, with the entry of a competitor, with the category's inflation and with the customer's memory. A measurement from a year ago describes a market that is no longer the same. It serves as a starting point and not as a fixed parameter — and that is why those who work with price measure again instead of keeping the number.
The terms that decide price together with this one
No pricing term explains itself alone. These are the ones that appear in the same decision, each with its definition and its calculation.
Further reading
What the blog has already published about price elasticity of demand and about what this definition addresses.
Understanding the calculation is one thing; redoing it across thousands of items is another
For one product, any of these calculations fits in a spreadsheet. Across an entire catalog, with cost changing on every incoming invoice, it stops being done — and the price ends up inherited from the previous month.