Test Your Discount Codes Before You Launch

Promotion logic fails in ways that look like nothing is wrong. The code applies, the order completes, and the margin is gone because two discounts stacked on a product that was meant to be excluded.

Nobody finds this on Black Friday. They find it in January, in the margin report. With the offer locking this week ahead of 27 November, the test matrix below takes an afternoon and is the cheapest margin protection available.

Why does promotion logic break?

Because a peak offer is several rules interacting, and the rules were written separately.

A sitewide code, a category promotion, a free shipping threshold, a bundle price, a loyalty discount and an influencer code all exist independently. Each one works. The combinations are what nobody tested, and during peak a shopper will find the combination you missed, usually within hours of launch because deal communities look for exactly this.

The failure is quiet. Orders complete, revenue looks strong, and the discount rate in your reporting sits well above what you modelled.

What should you test this week?

Ten cases, each as a real order through to the confirmation page rather than as a theory.

A single code on a full price item. A code on an already reduced item. Two codes attempted together. A code plus the free shipping threshold. A code on a bundle. A code on an excluded product or collection. A code below the minimum spend. An expired code. A code on a gift card. And a code applied to a basket that mixes included and excluded items.

The last one is the case that most often fails, because the rule was written at product level and the basket is where it gets applied.

What rules should be explicit?

Four, written down rather than assumed, because your team will be asked about all of them during the weekend.

Whether codes stack with each other. Whether codes apply to already discounted items. What the exclusions are, named by collection rather than described loosely. And whether the promotion runs automatically or needs a code entered.

Automatic discounts convert better than codes a shopper has to remember and type, which is worth knowing before you choose. Removing that friction is one of the cheapest conversion levers in the window, since it costs no margin at all.

What breaks at the basket rather than the product?

Thresholds and proportional allocation.

If a percentage discount applies across a mixed basket, check how the platform allocates it, because the split decides what a return refunds. A customer who returns the full price item from a discounted basket can end up refunded more than they effectively paid for it, which turns a returns spike into a margin event.

Check the free shipping threshold against the post discount total too. Most brands intend the threshold to apply before the discount and configure it after, which gives away shipping on baskets that were never meant to qualify.

How should the offer end?

On a stated date and time, configured rather than manual, with a decision already made about extensions.

A code that technically still works after the advertised end is a reliable source of chargebacks and complaints, and the screenshot circulates. Set the expiry in the platform rather than relying on someone switching it off at midnight.

Decide the extension question now as well. Whether the offer stops, tapers or rolls into a December promotion is a margin decision, and it should be made while the numbers are calm rather than at 9pm on Cyber Monday. The December handover is covered in the two weeks after Cyber 5.

What about influencer and affiliate codes?

They need separate rules, because they are the codes most likely to leak.

A creator code posted publicly becomes a sitewide discount within hours, stacked on top of whatever else is running. Set them as single use where the platform allows it, exclude them from stacking, and cap the discount depth lower than your headline offer rather than matching it.

The paid search side of the same leak, where affiliates bid on your brand plus discount code, is covered in brand defence.

What should support be told?

The answers before the questions arrive, because promotion queries are the largest pre purchase category during peak.

Write down what the offer covers, what it excludes, whether it stacks, what happens if someone bought the day before, and who can authorise a goodwill adjustment. The wider preparation is in customer service at peak.

The offer modelling this testing protects sits in the Q4 Profit Playbook, free and written for founder led DTC brands on Shopify at $5M to $30M. Webtopia runs peak trading for founder led brands as an ecommerce marketing agency.

Frequently asked questions

How do I stop discount codes stacking?

Set the combination rules explicitly in your platform rather than relying on defaults, then place a real test order attempting two codes together to confirm the rule holds at the basket.

Should Black Friday discounts be automatic or code based?

Automatic generally converts better, because a code the shopper has to remember and type adds friction at the point of purchase and costs no margin to remove.

What is the most commonly missed promotion test?

A basket mixing included and excluded items. The rule is usually written at product level and applied at basket level, which is where it fails.

How should influencer discount codes be configured?

Single use where possible, excluded from stacking, and capped below your headline offer, since public codes spread quickly and otherwise become a sitewide discount.

When should the promotion expiry be set?

In the platform, before launch, on a stated date and time. A code still working after the advertised end produces complaints and chargebacks.