You ran twenty percent off for a week. Orders came in thirty percent above a normal week, the revenue chart did the thing you wanted it to do, and everybody agreed the promo worked. Then the month closed and the money was not there.
The gap is not a mystery and it is not accounting. It is that the promo has to clear a bar almost nobody calculates before running it, and the bar is much higher than a thirty percent lift.
What the discount actually costs
Take a hundred dollar product that costs forty dollars to make, pick and ship. Sixty percent margin, which is a healthy consumer business. Sell it at full price and you keep sixty dollars.
Now take twenty percent off. You collect eighty dollars, the forty dollars of cost has not moved, and you keep forty. The discount was twenty percent of the price and a third of the profit. To end the promo with the same gross profit you started with, you need fifty percent more orders, not twenty percent more.
The general form is worth writing on something permanent. The extra volume you need is your margin divided by your margin minus the discount. Sixty percent margin and a twenty point discount gives 60 over 40, so 1.5x. At forty percent margin the same discount gives 40 over 20, which is 2x. Thin margins do not get to run sales, they just get to lose money faster.
Then you add the ad spend
That math assumed the orders arrive for free. They do not. Put twenty five dollars of acquisition cost on each order, which is modest for paid social, and the picture changes completely.
At full price you keep sixty dollars of gross profit and thirty five dollars after the ad cost. At twenty percent off you keep forty and fifteen. The number that pays for your rent, your team and your software just fell by more than half, on every order, including all the orders you were going to get anyway.
Here is what each discount level needs, at those numbers, to leave you exactly where you started:
- 10% off needs 40% more orders
- 15% off needs 75% more orders
- 20% off needs 133% more orders
- 25% off needs 250% more orders
- 30% off needs 600% more orders
- 35% off needs an infinite number of orders, because at that point the order clears exactly what it cost to acquire
That last line is the one worth sitting with. Somewhere below the discount most brands run at least twice a year, the business stops having a break even point at all. Every additional order makes the number worse. There is no volume that fixes it, and volume is the exact thing a sale is good at producing.
Read it the other way round
The useful version of this is not "how many orders do I need", because you find that out afterwards. It is "what discount can my realistic lift pay for", which you can answer before you commit.
At the numbers above, a fifteen percent lift in orders pays for about a five percent discount. A thirty percent lift pays for about eight percent. Doubling your orders, which almost never happens outside a launch, pays for about seventeen and a half percent.
So the twenty percent off that lifted orders thirty percent did not underperform. It performed well, and it still ended the week with forty four percent less contribution than doing nothing at all. It was a good promo attached to a number that could not work.
Free shipping is the same conversation wearing a costume. Twelve dollars of shipping absorbed on a hundred dollar order is a twelve percent discount, and it needs about half again as many orders to stand still. It tests better than a percentage off because it does not read as a price cut, but your margin cannot tell the difference.
Most of the discount goes to people who were already buying
The part that makes all of this worse is structural. The lift lands on the margin and the discount lands on the base.
If a hundred people were going to buy this week at full price and the promo brings in thirty more, you did not discount thirty orders. You discounted a hundred and thirty. A hundred of those people had their wallet out. The promo's real job was to be worth more than the money you handed back to customers who needed nothing handed back to them, and at a thirty percent lift and twenty percent off, it is not close.
This is also why redemption rate is the wrong number to celebrate. A code that gets used by eighty percent of buyers is not a sign the offer was compelling. It is a measure of how easy the code was to find.
How to find out what yours is worth
You can settle this inside one promo, with no extra spend.
Hold out a random slice. Take ten or twenty percent of the list at random and do not send them the code. Send them the same campaign without it. Then compare revenue per person, not open rate and not redemption rate. The difference between the two groups is the entire honest value of the discount, and it is usually a lot smaller than the sale week looks.
Compare on contribution, not revenue. Price minus cost of goods minus acquisition cost, times orders. Revenue always goes up during a sale. That is what a sale does. It is not evidence of anything.
Fence it so the base does not get it. New customers only, first order only, minimum cart value, a specific slow moving SKU. Every fence moves the discount off the people who were buying anyway and onto the people it was supposed to convince.
Try the offers that cost you less than they are worth. A gift with purchase costs you your cost, not your price, so it lands as a forty dollar gesture that costs sixteen. A bundle raises the cart while lowering the per unit price. A shipping threshold buys you order value instead of giving away margin. These are not clever tricks, they are just cheaper ways to say the same thing to the customer.
Watch what it does to next month. The most expensive part of a recurring promo is not the promo. It is teaching a list that full price is optional, which shows up as a quiet slump in the weeks between sales and never gets attributed to the sale that caused it.
Where we land on this
Every number above is arithmetic, not a benchmark, so put your own in. Your contribution per order before the discount, divided by your contribution per order after it, is the volume multiplier the promo has to beat. If that multiplier is bigger than any lift you have ever actually produced, the promo is a decision to buy revenue with profit, which is a legitimate thing to do deliberately and a bad thing to do by accident.
There is a version of this that we care about for obvious reasons. A discount is one way to squeeze more out of traffic you already paid for, and it is the expensive one, because it pays every buyer to do what most of them were going to do. The cheaper version is to get a second chance at the people who left without buying, and follow up with them at full price through the email you are already sending. Same traffic, same campaigns, no margin handed back. Whether that clears at your numbers depends on your traffic and your cost per order, and we are glad to run it with you and tell you if it does not.