Ecommerce discount strategy, done properly

Revenue is soft. The month is two thirds gone and the number is not where it should be. You need something you can run this week, and there is exactly one lever that works that fast.

So you run a sale.

The orders arrive. The spike is real, and it is measurable, and it happens within hours rather than months. Almost nothing else in a small ecommerce business responds that quickly, which is precisely why discounting is the tool most people reach for and the one they reach for hardest when they are worried.

This guide is not an argument against ever discounting. Discounts have a legitimate place, and there are four situations further down where running one is the right commercial call. The argument is against discounting by default: reaching for it because it is reachable, without measuring what it costs, what it teaches, or whether it worked.

The first cost: the margin, and it is bigger than it looks

Here is the thing that catches most people out.

A 20 percent discount does not cost you 20 percent. It costs you 20 percent of the price, and the price is not what you keep.

Take a product that sells for £40 with a 50 percent gross margin. It costs you £20, so you make £20 a unit. Take 20 percent off and it sells for £32. It still costs you £20. You now make £12.

You gave away 20 percent of the price and 40 percent of the profit.

The relationship is simple enough to do in your head:

Profit lost = Discount / Gross margin

At a 50 percent margin, a 20 percent discount costs you 40 percent of your profit. At a 30 percent margin, the same 20 percent discount costs you 67 percent. At a 25 percent margin, a 25 percent discount costs you all of it, and you are working for the practice.

The thinner your margin, the more violently this behaves. And most people run the same promotional depth across a catalogue where margins vary by twenty points, which means some of those products are being sold at a loss without anyone deciding that they should be.

What the promotion has to do to pay for itself

The profit drop is only half the picture. The other half is the question worth asking before you press send: how many extra units does this need to sell just to stand still?

Extra units needed = Margin / (Margin - Discount)

At a 50 percent margin with a 20 percent discount, that is 50 divided by 30, which is 1.67. You need to sell 67 percent more units to make the same profit you would have made anyway.

Not 20 percent more. Not “a decent bump”. Two thirds more volume, from the same audience, in the same week, before the promotion has earned back what it gave away.

It does not scale gently either. On that same 50 percent margin, a 10 percent discount needs 25 percent more units, a 20 percent discount needs 67 percent more, and a 30 percent discount needs 150 percent more. Each additional point comes out of a thinner slice than the one before it, so the requirement accelerates.

The full break-even method is here, including what to do with the number once you have it.

Most promotions do not hit their break-even. They look like a win on the revenue line and a loss on the profit line, which is easy to miss if the dashboard you check every morning shows you the first one.

The second cost: what it teaches your customers

Customers are pattern recognition machines. They are not studying you, but they notice, and they adjust.

Run a sale every November and your customers learn to wait for November. Send a discount code every time someone abandons a basket and your customers learn to abandon their basket. Offer a reduced rate every time someone tries to cancel and your customers learn that the way to get a better price is to threaten to leave.

None of those customers were retained. They were rented, at a lower price, and the moment you stop paying the rent they go.

There is a worse version of this. Discount often enough and customers stop believing the full price. If a product is available for 20 percent off most months, then 20 percent off is not a discount, it is the price, and the number on the page the rest of the time is a fiction that makes every full-price purchase feel like being caught out.

That perception is very hard to reverse once it sets, and it is invisible until you try to stop. How this pattern forms and how to unwind it is its own guide, because getting out is harder than getting in.

The third cost: who it brings you

A customer who came primarily for the discount is, by definition, price motivated. That is not a criticism of them. It is an observation about fit.

They are more likely to leave when the discount ends, more likely to move to whoever is cheaper next quarter, and less likely to become the sort of long term customer the rest of your retention work is built around. Discounting fills the top of the funnel quickly. It fills it with people who are harder to keep.

This is measurable in your own data, and you should measure it rather than take my word for it. Split last year’s customers into those whose first order used a discount code and those who paid full price, then compare repeat purchase rate and lifetime value across the two groups. If the discounted cohort is worth meaningfully less, your acquisition offer is not acquiring customers. It is buying orders.

The LTV calculator will give you the figure for each group, and customer lifetime value, explained properly covers why that number decides what you can afford to pay for a customer in the first place.

When a discount is genuinely the right call

Four situations. If your promotion does not fit one of them, it probably wants a second look.

A first purchase offer, where the discount is explicitly a trial mechanic. You are lowering the barrier to a first order in exchange for the chance to demonstrate what you are actually like. This only works if your retention is strong enough that people who arrive on a discounted first order go on to buy at full price. If they do not, you have not bought a customer, you have bought an order at a discount, and you will need to buy the next one too.

An infrequent promotional event, measured against baseline. Black Friday, end of season, an anniversary. These work when they are rare enough to feel genuinely unusual, when they sit on products carrying enough margin to absorb the cut, and when you compare what you sold to what you would have sold anyway. If you cannot show the promotion drove sales that would not otherwise have happened, it was not a promotion. It was a margin reduction with a countdown timer on it. Whether to run a Black Friday sale works through that comparison properly.

A targeted retention offer, to one customer, once. A subscriber who is about to cancel and has never had a discount before is a reasonable candidate for a one time reduced rate. The word doing the work is targeted. A blanket offer to everyone who contacts support about cancelling trains all of them to contact support when they want a better price.

Genuine overstock clearance. Moving old inventory at a reduced margin is often simply correct. Just be honest in your own accounting about what you are doing, and do not let clearance pricing quietly become a permanent feature of your emails.

What to do when the answer is no

Here is the awkward part. Most of the time you are reaching for a discount, the underlying problem is not price.

Revenue is soft because too few of last year’s customers came back, or because the first thirty days after a first order are silent, or because the people who used to buy every eight weeks have drifted to twelve and nobody noticed. A promotion does not fix any of those. It borrows a few weeks of demand from next month and charges you margin for the loan.

The alternatives are less immediate and they work better. A free bonus item in the first box usually costs less than a 25 percent margin reduction and creates a better first impression. A pause option for subscribers addresses the most common reason people cancel, which is a short term budget squeeze, without establishing a price expectation. A win-back email to customers who bought twice and then went quiet costs nothing and reaches people who already know they like the product.

What to run instead of a discount goes through the options and what each one actually costs you.

The rule

Three lines, and the difficulty is in following them rather than remembering them.

Never discount without measuring whether it drove behaviour that would not have happened at full price. Never discount the same customer repeatedly, because repetition becomes expectation. Never discount as a substitute for the relationship work that would make discounting unnecessary.

Discounts are a tool. Like every tool they work when used deliberately and cause damage when used by default. Know why you are reaching for this one before you do, and run the number first.