When to Retire a Digital Product (and How to Take It Down Properly)

Published July 26, 2026

Part of: Digital Products — our full guide on this topic.

Disclosure: Some links below are affiliate links. If you sign up through them we may earn a commission at no extra cost to you. We only recommend tools we'd genuinely suggest to a friend. See our full disclosure.

Every guide on this site takes a digital product forward. Build it, price it, launch it, update it, and when it is quiet, run a promotion or relaunch it.

Nothing tells you when to stop.

So the default, for almost everybody, is that nothing is ever stopped. Products accumulate. The shop page gets longer. Two or three things sell, the rest sit there, and each one is a small unresolved decision you walk past every time you open the dashboard. Eventually you have a catalogue you would not confidently recommend as a whole, and no process at all for the awkward question underneath it: is this thing still worth selling?

A product that is not selling is not free to keep — but it does not cost what people assume, and “no sales” is a symptom with four different causes, only one of which is the product’s fault.

That is the first half of this guide: how to tell a quiet product from a dead one, so you retire the right things and stop retiring your own inattention.

The second half is the part nobody plans for. Taking a product off sale and taking it away from the people who already paid for it are two different actions — and you can very easily do the second while intending only the first. Your buyers are holding a copy of the file from the day they bought, and a surprising number of them are not holding it at all: they are relying on a receipt link they assume will keep working. What happens to that link when the listing comes down is not something to guess at.

Honest disclosure: some links below are affiliate links. If you sign up through one I may earn a commission at no extra cost to you. Everything here is my genuine assessment.

The quick version

• Quiet and dead are different. A product that sells three times a year and needs nothing from you is a working asset, not a failure.
• Diagnose before you decide. Nobody saw it, they saw it and did not click, they clicked and did not buy, or the problem it solves has gone. Only the last one is a product problem.
• Never retire something you have not honestly promoted in the last six months. That is not a verdict on the product, it is a verdict on you.
• There are three exits, not one: freeze it, unlist it, or retire it. They differ in what they cost and in what they do to past buyers.
• Email your buyers before you take anything down, with a date and a fresh download link. One email removes the entire access question.
• Do not delete the URL. Redirect it or replace it with an honest retirement note — the page’s links and search visibility belong to the address, not the product.
• A retired product has three afterlives: free lead magnet, part of a bundle, or raw material for content. Deleting it is the only option that recovers nothing.

First: quiet is not the same as dead

Three completely different situations get filed under “it is not selling”, and they call for opposite responses.

It has never sold at all. No purchases since launch. This is the one that feels like a verdict and usually is not — see the next section, because in the overwhelming majority of cases the product was never actually put in front of enough people to have failed.

It sold, then stopped. There was a launch, or a period where it moved, and then it went flat. Something changed: a traffic source dried up, the page fell out of search, a competitor appeared, or the thing it was about moved on. This is the most informative case, because there was demand once, which means the product is not the problem — the change is.

It sells rarely but reliably. Two in March, one in July, one at Christmas. This is not failure. This is the long tail, and it is most of what a small catalogue’s income actually is. A finished product that still works and generates no support is close to free to keep on sale, and the temptation to “clean up” by removing it is a tidiness instinct pretending to be a business decision.

Being clear which of the three you have takes five minutes with your own dashboard and saves you from the most common mistake in this whole area: retiring the long tail because it looks unimpressive next to the bestseller.

The four reasons a product stops selling — only one is the product

“No sales” is not a diagnosis. It is a symptom at the end of a chain, and you can locate where the chain breaks by asking four questions in order. This matters because three of the four have nothing to do with the product itself, and retiring the product fixes none of them.

1. Nobody saw it. The sales page had almost no visitors. This is by far the most common answer and the most consistently ignored, because it is the least flattering. A page nobody visits cannot sell anything regardless of how good the thing behind it is. If this is your answer, the product has not been tested — traffic has. Do not retire it; promote it once, properly, and then judge.

2. They saw it and did not click. People reached the page or the listing and left. Now you are looking at the promise, not the product: the title, the first line, the images, whether the thing it claims to do is a thing anybody wants. That is a sales page and description problem, and it is cheap to test — you can rewrite a page this afternoon and you cannot rebuild the product this afternoon.

3. They clicked but did not buy. They got as far as checkout and stopped. Price, format, trust, or a missing piece of proof. This is where a review or two does more than any amount of copy, and where a genuinely mismatched price shows up.

4. The problem it solves has gone. The software it teaches was redesigned or retired. The platform it was built for shut down — and note that if the platform you sell on is the one closing, that is a different emergency entirely: the product is fine and the shelf is being removed, which is nothing like the decision on this page. The format is obsolete — a template for an app nobody uses, a guide to a feature that no longer exists, a checklist for a process that changed. Nobody is looking for this any more, and no amount of promotion changes that.

Only the fourth is a product problem. The first three are marketing problems wearing a product’s clothes, and retiring the product in response to them is expensive in a way that is easy to miss: you throw away finished work to avoid an afternoon’s rewriting.

And even the fourth has a fork. If the underlying need still exists and only the details have aged, that is not a retirement, it is an update — and it is worth being honest about which one you are choosing, because “it is out of date” is a much more comfortable reason to quit than “I do not want to maintain this any more”. Both are legitimate. Only one of them is what you will tell yourself.

What a dead listing actually costs you

The case for leaving everything up forever is that digital products have no storage cost, and that is true. But “no storage cost” is not the same as “free”, and there are four real costs worth naming — plus one that people invoke and should not.

It competes with your good products for a decision. A visitor looking at a shop page with fifteen items is not fifteen times more likely to buy than one looking at four. Choice past a certain point stops helping and starts producing hesitation, and the item that absorbs somebody’s attention before they give up was not free — it cost you the sale that the good product would have made.

It is a live obligation. As long as it is buyable, somebody can buy it, and the moment they do you owe them a working file, working links, and an answer to their support email. A product you have mentally abandoned but left on sale is a promise you have stopped intending to keep. That is the version of this that turns into a refund request and then a review you cannot answer well.

It concentrates your bad reviews. If one product in a catalogue is out of date, it is the one collecting the disappointed buyers — and on most platforms that feedback is attached to you, not only to it.

It is a public artifact with your name on it. Somebody encountering you for the first time may well encounter that — the abandoned one, the one with the screenshots from three redesigns ago — and form their view of everything else from it.

The cost people invoke and should not: platform or hosting fees. If you are on a plan at all, you are almost certainly paying for the plan and not per product, so an extra listing is not what is costing you money. If fees are genuinely the driver, the decision you are actually making is about which platform you sell on, not about which product you kill.

Notice what all four real costs have in common: none of them applies to a product that still works, still delivers correctly, and generates no support. That product costs you nothing, and it is the one you should stop feeling guilty about.

One boundary worth stating before going further: everything in this article assumes the business carries on around the product. That is what makes “email the buyers, keep serving the file, redirect the URL” possible — there is still a monitored inbox, a live shop and a working payment account behind those instructions. If what you actually want to retire is the whole thing, every one of those steps needs a different answer, and closing a business down is where they are. There is a second thing this article assumes and does not say aloud: that the product is finished, so the buyer already holds it and one last download link makes them whole. Nothing here transfers to a membership or a subscription, where what you are switching off is still being used and still being billed — closing a membership is the version of this with a live obligation attached.

The three exits, which are not the same action

“Retire it” gets used for three genuinely different things. They differ in effort, in reversibility, and — the part that matters most — in what they do to people who already paid.

Freeze it. Stop promoting it. Leave it on sale, exactly as it is. Remove it from your own content rotation and your emails, and let it sell to whoever finds it.

This is the right answer far more often than people expect, and it is the only one of the three with essentially no downside: no buyer is affected, no URL changes, the long tail keeps running, and you have simply stopped spending attention on it. The precondition is honest — the product must still work, and you must still be willing to answer an email about it. Freezing something broken is not freezing, it is abandoning it in public.

One genuinely good move here: put a dated scope line on the sales page. “This covers the 2026 version of X, as of March 2026.” A dated, accurate snapshot is a perfectly legitimate product. An undated one that is quietly six years old is not.

Unlist it. Take it off your shop page and out of your navigation, but leave the product page itself reachable by direct link.

This is the useful middle option and the one most people do not know they have. It is right when you want to keep honouring the thing without offering it to new visitors: you have folded it into a bundle and want the standalone version to stop competing, or you are keeping it available for a specific group, or you want to stop new sales while you decide. Existing buyers are typically unaffected because the page still exists — but this is exactly where you should verify rather than assume, and the buyer email in the next section makes the question moot anyway.

Retire it. Off sale, gone from the shop, no longer purchasable by anybody.

This is the only one of the three that is meaningfully hard to reverse, and the only one with a real obligation attached. Everything in the next two sections is about doing this one properly.

The practical rule: default to freeze, escalate to unlist when it is crowding something, and only retire when the product should genuinely not be sold to anyone else. Most catalogue cleanups that people describe as ruthless pruning would be better done as a freeze, because what they actually wanted was to stop thinking about it.

Before you take anything down: what you owe the buyers

This is the step that gets skipped, and it is the one with a person on the other end of it.

Your buyers are holding a copy of the file from the day they bought — that is the whole reason updating a digital product is awkward. But a good number of them are not holding it at all. They bought it, skimmed it, and left it in the receipt email, treating the download link as storage. That is a completely normal way for people to behave, and it means the day you take a listing down, you should assume at least one person’s only route to the thing they paid for goes through you.

What happens to an existing buyer’s download link when a listing comes down is not something to reason about from first principles. It differs between platforms, and it is not always documented where you would look first. So do not build your plan on an assumption in either direction. Build it on this instead, which works regardless:

Email everyone who bought it, before you take it down. Give them a date, and give them a link that works right now. That single email removes the whole question — if the link keeps working afterwards, fine; if it does not, everybody who cared already has the file.

The order of operations is: write the email → give a real window → then take it down. Not the other way around, and not the same day. A couple of weeks is enough for someone to see the email and act on it, and it is the difference between a courteous retirement and a customer discovering by accident that something they paid for has vanished.

Two specific cases deserve care. If you ever sold it with “lifetime updates” or anything similar, retiring the product is the end of that promise, and the honest thing is to say so plainly in the email rather than let it lapse silently. (This is precisely why “free updates for as long as this product is on sale” is a better sentence to put on a sales page than “lifetime updates” — it tells the truth about the fact that products get retired.) And if the product is part of a bundle someone bought, they did not buy the retirement of a component, so either keep serving them the file or replace it with something of equivalent value.

All of this depends on being able to reach the people who bought — which is a good moment to find out whether you can, and a good argument for keeping your own copy of the buyer records rather than discovering their format for the first time under time pressure. Most direct-checkout platforms record a buyer email even if you never thought of it as a list, and it can usually be exported or mailed from inside the tool. If your store and your email software are separate, there is an export-and-import step in the middle, and it is exactly the kind of friction that turns into “I will do it next week”.

That is the practical argument for keeping the store and the list on one account. Systeme.io is the one I point people to for this: products, buyer records and the email list live together, so “email everyone who bought this one thing” is a filter rather than a data migration, and there is a free tier a small catalogue runs on fine. (That is an affiliate link — if you upgrade to a paid plan through it I may earn a commission, at no extra cost to you; see the full affiliate disclosure.) Free tiers and features change, so confirm the current details on the provider’s own site. If you have no buyer list at all, collecting email addresses is the fix, and this is one of several moments where its absence costs you.

Marketplaces are the constrained case. On a platform that owns the customer relationship, what contact details you receive and what you are permitted to send both vary, so read that specific marketplace’s rules before writing to anyone rather than treating a buyer list as a mailing list.

The retirement email

Five lines. It is not a marketing email and it should not read like one.

• Subject: the product name, and that it is being retired. Nothing clever. It only has to be recognised by somebody who bought from you eighteen months ago.
• Remind them who you are and what they bought. “You picked up the X pack last spring.” Half of an old buyer list will not recognise your name on sight.
• The date it comes off sale, stated plainly.
• One link: download it now. This is the entire purpose of the email.
• One line on why, without drama. “I am no longer keeping it up to date, and I would rather retire it than sell something I am not maintaining.” That sentence does you credit and costs nothing.
• At most one line about anything else, and only if there is a genuinely relevant replacement. An email that exists to protect somebody’s purchase should not turn into a pitch.

If it had a “lifetime updates” style promise on it, add one honest sentence: this is the last version, and it is theirs to keep. People are far more understanding about a clearly-stated ending than about a promise that just quietly stops being true.

What to do with the URL

The most valuable thing about an old product page is often not the product. It is the address.

A page that has been live for a while may have accumulated search visibility, links from other sites, links from your own articles, saved bookmarks, and a Pinterest pin or two still driving traffic. None of that was earned by the file — it was earned by the URL, and it survives the product’s retirement unless you delete it.

So do not delete. Do one of these:

Redirect it to the nearest genuinely relevant live thing. The successor product, the category page, or the article that covers the same ground. “Genuinely relevant” is the operative constraint: a redirect to something unrelated is a worse experience than an honest dead end, because the visitor now has to work out what happened.

Or replace the page with a short retirement note. Three sentences: this product has been retired, here is why in one line, here is what I would recommend instead. This is often better than a redirect, because it answers the visitor’s actual question rather than silently teleporting them, and it keeps a real page at an address that other pages still point to.

Do not redirect it to your homepage. Somebody who wanted one specific thing and lands on a generic front page does not browse — they leave.

Worth noting that this rule is stricter for a product page than for an ordinary article, and for a specific reason: your buyers have a relationship with this exact address — a receipt, a bookmark, a download link they expect to keep working — so it has to keep saying something. A retired blog post has no such holder, which is why the rules for an ageing content archive allow an honest 404 where this does not.

And then fix your own house: search your own site for links to the retired product and update them. Your articles are the traffic source most likely to still be pointing at it, and an internal link to a dead product page is the one broken link that is entirely your own doing.

The three afterlives of a retired product

Retiring does not have to mean the work evaporates. A finished product you are no longer selling is still an asset, and there are three honest things to do with it. Deleting it is the only option that recovers nothing.

Turn it into a lead magnet. A product nobody will pay for can still be worth an email address, and this is the most consistently underrated move on the list. The work is done, the file exists, and the same thing that failed at £19 may be an excellent lead magnet — genuinely useful, specific, and immediately deliverable. One caveat that matters: if people paid for it recently, giving it away free the following week is a poor thing to do to them. Leave a decent gap, or offer those buyers something in return.

Fold it into a bundle. Two or three products that are individually too thin can be one product that is not. This is often the better answer for a group of small, related things that each sell rarely — bundling turns a cluttered shop page into a single stronger offer. Look after the people who bought the parts: a free upgrade to the bundle is cheap, generous, and exactly what you would want if you were them.

Break it up into content. A retired guide is several articles. A retired template pack is a tutorial. Repurposing it puts the work in front of people permanently, keeps the expertise visible, and can send traffic to whatever you sell now — which is more than it was doing sitting unbought in a shop.

The one thing to keep in all three cases: the editable source files. A product you can only ever republish as a flat export is a product you can never revive, split, or improve. Keep the design file, the document, the spreadsheet, and keep them somewhere you will still find them in two years.

When not to retire

Five situations where the instinct is wrong.

You have not actually promoted it. The single most common case. If the product has not had one deliberate promotion attempt in the last six months — a real email to your list, a month of content pointing at it, a genuine offer — then you are not retiring a failed product. You are retiring your own inattention, and you will learn nothing from it.

It is seasonal. Some products have one month a year. Judging a Christmas product in April and a tax-season product in September will retire perfectly healthy things.

It feeds something else. A cheap product whose real job is to turn a stranger into a buyer is not judged on its own revenue. If it is the entry point to your catalogue, its numbers are only half its contribution, and you cannot see the other half by looking at its sales line alone. This is the trap of judging products individually rather than looking at what the business as a whole is doing.

It sells slowly and costs nothing. Covered above and worth repeating, because the tidiness instinct is strong: slow is not a reason. Slow plus out-of-date, or slow plus support you resent, is a reason.

You are retiring it because you are bored of it. You have looked at it a hundred times more than any buyer has. Your fatigue with a product is not evidence about its value — it is evidence about how long you have been staring at it. This one is worth naming, because it disguises itself as judgement extremely well.

There is also a sixth situation where the answer is neither retire nor leave alone: the product is worth keeping but the shape of it is wrong — the format, the platform it lives on, what is bundled with it. That is not a retirement, it is a change made to something people have already paid for, and it comes with obligations to past buyers that taking a product off sale does not.

The decision, in order

1. Which is it — never sold, sold then stopped, or slow but steady? Slow but steady: stop here, leave it alone.
2. Has it had a real promotion attempt in the last six months? No: promote it once, properly, then come back.
3. Where does the chain break — seen, clicked, bought? Anywhere but “the problem has gone”: fix that instead of retiring.
4. Does it still work, and would I still support it? Yes: freeze it. That is a complete answer and it is free.
5. Is it crowding better products, or does it need to stop being sold? Unlist, or retire.
6. If retiring: email buyers with a date and a link, wait, then take it down.
7. Redirect or replace the URL. Never delete it.
8. Check where your bio links point. Social profiles you no longer post on are the most likely place a retired product’s old link is still live — a quick pass over them catches it.
9. Reread your email sequences. Anything automated that mentioned or sold this product is still sending, and unlike a web page you cannot fix the copies that have already gone out — you can only shorten how long the broken version keeps going. Auditing your email automations is the pass; retiring a product is one of the four events that should trigger it.
10. Pick an afterlife: lead magnet, bundle, or content. Keep the source files.

The bottom line

The reason nobody writes about this is that retiring a product feels like an admission, and launching one feels like progress. But a catalogue is not a collection of everything you have ever made — it is what you currently stand behind, and keeping it that way is ordinary maintenance, not defeat.

Diagnose before you decide, because three of the four reasons a product stops selling are not the product’s fault. Default to freezing rather than removing, because it costs nothing and affects nobody. When you do take something down, tell the people who paid for it first, with a date and a working link, and never make somebody discover by accident that a thing they bought has disappeared. Keep the URL working. Keep the source files. And judge the thing on whether it has actually been given a chance, not on how tired of it you are.

Then go and put the attention you just freed up into the products that are working — which is the entire point of retiring anything.