The code graveyard: a quarterly clean-up for old discount codes

An order comes in and the total looks wrong by about a quarter. You check, and the buyer used a code you wrote for a webinar in November. You have not mentioned it since December. They did not steal anything — it was live, they typed it, it worked. Somewhere between a coupon extension, a deal forum and an old email sitting in a public archive, a code you had mentally retired eight months ago has been quietly doing business on your behalf at a price you no longer agree with.

Why old codes stay alive

Every campaign has a natural end in your head and no natural end anywhere else. You stop sending the email, so the campaign feels over. But a code is a string of characters, and strings of characters travel. Someone screenshots your story. Someone posts it in a group chat because they are being helpful. A coupon extension scrapes it and offers it automatically to anyone who reaches your checkout, forever, with no way for you to ask it to stop.

The honest difficulty is that you often want some codes to live a long time. A partner code you gave to a podcast host is meant to keep working, because their back catalogue keeps playing. A welcome code inside your automated sequence should not die because you did a tidy-up on a Sunday. So a clean-up cannot be a blanket delete. It has to be a decision, code by code, about which ones you still mean.

It helps to know where the leaks actually are, because they are boringly predictable.

• Video descriptions and pinned comments, which nobody ever goes back to edit.
• Podcast episodes, where the read is baked into the audio and cannot be changed at all.
• Deal sites and coupon browser extensions, which will keep suggesting a working code long after you forget it exists.
• Newsletter archives and forwarded emails sitting in inboxes you will never see.
• Your own link in bio, if you added the code to a card during a launch and never took it down.

Leaving the list alone, versus twenty minutes a quarter

This is not a store.fan problem versus somebody else's problem. It is the difference between a codes list that reflects your intentions and one that reflects your history. Discount codes are a Pro feature, and like every powerful setting they reward being reviewed on a schedule rather than in a panic.

The quarterly pass

Put it in the calendar for the first working day of January, April, July and October. It is a twenty-minute job if you do it four times a year and a two-hour job if you do it once. Work through the list top to bottom and give every code one of three verdicts: keep, retire, or replace.

What the leak is worth

Suppose one forgotten code at 25% sat live for nine months against a £48 course, and twenty-two people used it. That is £264 of margin that left without a campaign attached to it, which is roughly a fortnight of the kind of work that actually grows a store. The number is small enough to ignore every single month and large enough to matter across a year, which is exactly why it survives.

The second cost is subtler and worse. Those twenty-two people are now in your customer list with a lower lifetime value than they would otherwise have, and your read on what that course is worth is quietly distorted. When you sit down in six months to decide whether to raise the price, you will be looking at averages that include a discount you did not mean to give. These figures are an illustration of the shape of the problem, not a claim about your store.

The mistake most people make

The mistake is editing instead of retiring. You find an old code at 30%, decide 30% is too generous, and change it to 15% so the link in that video still works. Now the code called AUTUMN30 gives 15%, and the next person who reads your list learns something false. Worse, somebody arrives from that video expecting 30% and gets a smaller discount with no explanation, which feels to them like a bait and switch even though you were trying to be accommodating. Retire it, make a new one, and let the old link fail cleanly. A code that does not work is a much smaller disappointment than a code that works differently than promised.

Four twenty-minute passes a year is less time than you will spend rewriting one launch email, and it is the difference between a price list you control and one that has been edited by strangers on the internet. Put the next pass in the diary before you close this tab, and if you are unsure how any of the settings behave, the FAQ is the fastest place to check.