Go back through your store, season by season, and count the products you have added since you opened.
Now count the ones you have taken away.
For most owners the second number is zero, or close enough to it that the exceptions are memorable. A supplier stopped making something. A line got recalled. Beyond that, the catalog has only ever gone in one direction, because adding is a decision somebody makes and removing is a decision nobody is assigned.
That is not carelessness. Launching a product has an owner, a date, a budget and a photographer. Retiring one has no meeting, no calendar entry and no obvious moment when it is supposed to happen. So it never happens, and the catalog quietly fills up with things that are still listed, still photographed, still occupying a slot in your navigation, and no longer paying for any of it.
You Cannot Cut What Your Store Will Not Show You
There is a reason this particular pile never gets dealt with, and it is more interesting than laziness.
Open your store's reporting and look at how products are ranked. They are ranked by revenue, or by units sold, because those are the numbers the platform can produce without knowing anything about your business. Your store knows what it charged. It does not know what you paid for freight in March, what the duty came to, what your fulfillment partner charges to pick a heavy item, or which of your products gets returned twice as often as the rest.
So the report you have answers "what sold." The question you need answered is "what paid." Those are different questions, and for a meaningful slice of most catalogs they give different answers.
This is the whole reason the tail survives. Nobody is defending the dead products. They are simply invisible in the only ranking anyone ever looks at, and a thing you cannot see does not get cut. We have written before about the gap between your best seller and your most profitable product, which is the same measurement problem pointed at a single item. This is what it looks like pointed at the whole catalog at once.
The Bill Nobody Sends You
The reason to care is that a product does not sit in your catalog for free. It just gets billed to accounts nobody reconciles against it.
The first and largest is cash. Every unit sitting in your stockroom is money you already spent, waiting to become money again, and a product that turns twice a year holds that money six times longer than one that turns monthly. That capital is not available for the product that is actually working. We looked at the mechanics of this recently in the piece on how your profit ends up sitting on a shelf instead of in your bank account, and the slow tail of a catalog is where that effect concentrates.
Then there is everything the product consumes without ever being charged for it. Photography and copy for each new variant. A slot in your navigation and your collection pages, which is genuinely finite, because a shopper's patience for scrolling is finite. Ad budget, if the product sits in a shopping feed or a catalog campaign, where it will happily spend against clicks it cannot convert profitably. Support time, because the products people are confused by are rarely the ones that sell well. And your own attention at every reorder, every stocktake and every promotional planning session, spread thinner across items that do not deserve an equal share of it.
None of those arrive as an invoice with the product's name on it. That is exactly why the tail grows.
Two Numbers Decide It, and Revenue Is Not One of Them
You need two figures per product, and neither is complicated.
The first is contribution margin per unit. Take what the product actually sells for after discounts, then subtract what it actually costs you to land and deliver: the unit cost from your supplier, freight and duty, packaging, the pick and pack fee, the payment processing on the sale, and an allowance for your return rate on that item. What is left is the money that product contributes toward your overhead and your profit.
Notice what this is not. It is not gross margin, which is the number most owners quote from memory, and which hides shipping and fulfillment entirely. And it is certainly not revenue, because revenue rewards the products you discount hardest.
There is a wrinkle worth knowing. That landed cost is not a permanent fact you calculate once and file. Your supplier's price moves when they renegotiate. Freight moves seasonally and by route. Duty moves when trade policy moves. If you buy in a currency you do not sell in, the exchange rate moves daily. A margin figure is a snapshot with a date on it, and a number you worked out eighteen months ago is describing a business that no longer exists.
The second figure is inventory turns: how many times a year you sell through and replace your holding of that item. Units sold in a year divided by the average number of units you hold. You do not need it to a decimal. You need to know the difference between a product that turns eight times and one that turns once.
Take those two numbers for your twenty biggest products and the argument usually settles itself. The item ranked second by revenue turns out to be seventh by contribution, because it is heavy to ship and gets returned. The quiet item in eighth place turns out to be the best margin in the catalog, and you have never once featured it.
Four Quadrants, Four Different Jobs
Put margin on one axis and turns on the other and every product in your catalog lands somewhere useful.
High margin and fast turning is where your advertising, your homepage and your email calendar should already be pointing, and often is not. Low margin but fast turning is a keeper, usually a volume item that brings people in and earns its place even though it contributes little per unit.
High margin and slow turning is the interesting quadrant, because it is nearly always a marketing failure rather than a product failure. The economics are good. Nobody knows it exists. That is a merchandising and campaign problem, and it is the cheapest growth available to most stores, which is the kind of work we do for ecommerce clients before anyone suggests increasing the ad budget.
Low margin and slow turning is the kill list. And it is almost always smaller than owners fear, which is the reassuring part of doing this exercise at all.
Three Questions Before You Kill Anything
The quadrant gets you a shortlist, not a verdict. Three questions separate the genuinely dead from the merely neglected.
Does it bring you customers who then buy something else? Some products earn their place by being the first thing a new customer is willing to try. If the item is unprofitable on its own but consistently shows up as somebody's first order before a profitable second one, you are looking at an acquisition cost wearing a product's clothing. Check the follow-on behavior before you cut it.
Is it doing a job for the product next to it? A high-priced item that rarely sells can still be making the item beside it look like the sensible choice. Remove the expensive one and the mid-priced one becomes your most expensive product, which changes how it reads. That is worth knowing before you delete it.
Is the problem the product, or the price? This is the one that rescues most of the list. If your costs have moved and your price has not, you do not have a dead product. You have an old price. Raising it is faster, cheaper and less disruptive than a discontinuation, and it frequently moves the item straight out of the kill quadrant.
How to Retire a Product Without Burning the People Who Bought It
Once something is genuinely going, the way you end it matters more than owners expect, because the people most affected are your repeat customers.
Tell them before they find out on their own. A short, unapologetic note to anyone who has bought the item, saying it is being discontinued and this is the last of the stock, is both a courtesy and usually the best sell-through you will get. People who love a product buy two when they hear it is ending.
Sell through the remaining units without teaching a habit. There is a real tension here, because the obvious move is a deep discount, and a predictable discount cadence trains your customers to wait for the next one. We made that argument at length in the piece on why a discount is not an offer. A final-units clearance is one of the few genuinely legitimate uses of a markdown, precisely because it cannot repeat. Say plainly that it is ending, and let the ending be the reason rather than inventing a sale.
Then deal with the page. Do not delete the URL. A discontinued product page that has accumulated links and search traffic for three years should redirect to the closest current alternative, or to its collection, so that the traffic and the trust land somewhere useful instead of on a 404. Pull the item from your shopping feed and any catalog campaigns on the same day, because an ad still running for something you no longer sell is the most expensive kind of dead stock.
And leave a note of what you learned. A product that failed because the freight cost doubled is a different lesson from one that failed because nobody wanted it, and in a year you will not remember which was which.
Every Product Needs a Review Date and Someone Who Can Say No
The reason this problem returns is that the fix is usually run as a one-time cleanup rather than as a habit.
The simplest change is to give every product a review date at the moment you launch it. Not a kill date, which sounds gloomy and gets ignored. A date, six or twelve months out, when somebody looks at its margin and its turns and decides whether it stays. It costs nothing to set and it converts a decision nobody owns into one that appears on a calendar. This is the far end of the same lifecycle we covered in the piece on launching a product properly rather than just adding it to the site: a launch with a review date is a plan, and a launch without one is an addition to a pile.
The second part is less comfortable. A review is worthless if nobody in the room is allowed to end anything. If the product stays because it was somebody's idea, or because you have eleven cartons of it in the back and removing it would mean admitting the eleven cartons were a mistake, then what you are holding is not a review. The eleven cartons are already spent either way. The only live question is whether the product gets any more of your money, your page space and your attention.
Start With the Bottom Ten
You do not need a system for this, and you certainly do not need software before you have done it once by hand.
Take your ten slowest-moving products. Work out contribution margin per unit and rough annual turns for each, using real landed costs rather than the numbers you carry in your head. Sort them. Then run the three questions against the worst four or five.
Most owners who do this find one or two obvious removals, a couple of items that just need a price correction, and at least one genuine surprise: a product with excellent economics that has never been promoted because it was never near the top of a revenue report.
That last one is usually worth more than everything you cut.
If you would rather have someone else do the arithmetic and tell you what it means for your advertising, that is a large part of what we do for store owners. Either way, the exercise is worth an afternoon. The catalog has been growing in one direction for years, and nothing is going to reverse that except somebody deciding to look.
