Use sell-through and days of stock cover to find real overstock, give it exposure before a markdown, avoid the made-to-order trap, and order your Shopify clearance collection by what clears at a profit.
Sell-through and days of stock cover are the two numbers that should decide your clearance merchandising on Shopify, and most stores use neither. Instead, when the stockroom fills up, the reflex is a blanket markdown: 30% off everything in the category, a banner, an email. Stock moves. So does margin, including on products that would have sold at full price if anyone had been able to find them. This post is about doing clearance properly: finding the products that are genuinely overstocked, giving them exposure before you give them a discount, avoiding the made-to-order trap that fools most stock logic, and ordering your Shopify clearance collection so it clears at the best margin available.
It is written for merchants who hold physical stock and feel it: apparel, homeware, beauty, sports, gifts. We explain each metric, show the sequence we would run, and say where SellRank, our Shopify collection sort app, automates each step. Example numbers are illustrative, worked through to show the logic.
Why blanket discounting is the expensive way to clear stock
A blanket markdown has three costs that do not show up on the day it runs. First, it gives margin away on products that were not overstocked at all, because a category-wide discount cannot tell a slow mover from a steady seller. Second, it trains customers: shoppers who learn that your store runs 30% off every few months stop buying at full price in between. Third, it hides the real problem. Many overstocked products are not unwanted; they are unseen, sitting on page three of a collection sorted by lifetime sales, and a discount fixes the symptom while leaving the ordering that caused it untouched.
Illustrative example: a homeware store has 400 units of a ceramic vase that sells about four a week. On a 'Best selling' ordered 'Vases' collection it sits at position 38, below the fold on every device. A 25% markdown moves some units. Moving the same vase into the second row for three weeks at full price might move a similar number, at no margin cost at all, and if it does not, you have learned that the problem is the product rather than the placement, and the discount is now justified.
What sell-through actually measures
Sell-through is the share of available stock that has sold over a period. A buyer typically computes it as units sold divided by units received, over a season. For collection merchandising, the useful form compares recent units sold with units still in stock: a product that sold 50 units in the window and still holds 50 has sold through half of what it had available; one that sold 5 and holds 95 has sold through a twentieth.
Sell-through is powerful because it is relative. A product selling 10 units a week sounds healthy until you see it has 800 units on hand; a product selling 2 a week sounds weak until you see it has 4 left. Raw orders and revenue cannot tell those two stories apart. Sell-through can.
- High sell-through, low stock: a winner that is running out. Stop pushing it hard, or reorder.
- High sell-through, healthy stock: a winner with depth. This is what should lead your main collections.
- Low sell-through, low stock: a tail product. Leave it where it is; it will clear itself or not matter.
- Low sell-through, high stock: genuine overstock. This is where clearance merchandising should focus.
One caution on reading sell-through: it is only as fair as the exposure behind it. A product that has spent the whole window below the second scroll of its collection has had a fraction of the views of the products above it, so its low sell-through is partly a measure of where you put it. That is exactly why the sequence below starts with exposure rather than price. Treat a low sell-through figure as a question to test, not as a verdict, until the product has had a fair chance to be seen.
What days of stock cover tells you that sell-through does not
Days of stock cover, sometimes called days of inventory or weeks of cover, estimates how long current stock will last at the recent rate of sale. If a product holds 120 units and has been selling 4 a day, it has about 30 days of cover. If it holds 120 and sells 1 a week, it has more than two years.
Cover turns stock into time, which is how clearance decisions are actually made. You rarely care that you have 300 units; you care whether they will be gone before the season ends, before the storage bill arrives, or before the next version of the product lands. A product with 400 days of cover in a seasonal category is a problem today, even if its absolute stock level looks unremarkable.
Cover also works at the other end. A best-seller with five days of cover and no reorder coming should start moving down the page before it sells out, so that it does not break at the top of your busiest collection. That is the same Shopify sort by inventory logic that keeps sold-out products out of the first row, applied a week earlier.
Using sell-through and days of stock cover together to find real overstock
Neither number alone is enough. A new product will show low sell-through simply because it has not had time to sell. A long-standing tail product can have huge cover on tiny stock that does not matter. The combination is what isolates the problem products:
- Filter to products that have been live long enough to have a fair rate of sale, for most stores at least four to six weeks.
- Look for low sell-through and high days of cover together. These are the products where stock is large relative to demand, and demand is not catching up.
- Weight by what the stock is worth. Two hundred units of a $6 item and two hundred of a $120 item are not the same problem.
- Sanity-check against visibility. Before calling something unwanted, check where it sits in its collections. Low sell-through at position 70 is weak evidence of anything.
SellRank offers Sell-through and days of stock cover as ranking strategies on every plan, including Free. That means you can open any collection in the preview, rank it by cover, and see the overstock list for that collection, product by product, before deciding anything.
The made-to-order trap
Here is where most stock logic goes wrong. Many Shopify stores sell a mix of physically stocked products and products that are made, printed or sourced to order. Merchants handle these in different ways: inventory tracking switched off, 'continue selling when out of stock' switched on, or a nominal quantity entered so the product shows as available. Each of these produces nonsense in naive clearance maths.
- Tracking switched off: the product has no stock figure, so a tool may treat it as zero stock and bury it as sold out, or treat it as unknown and rank it unpredictably.
- Continue selling with zero or negative stock: a stock cover calculation reads zero cover, and the product looks like a winner about to run out.
- A nominal large quantity, such as 999, entered so the product always shows as available: the cover calculation reads years of stock, and the product lands at the top of every clearance slot as the worst overstock in the store.
The third case is the one that hurts. A made-to-order ring with 999 nominal units and a steady rate of sale looks, to a naive tool, like the most overstocked product you own. Push it into a clearance slot, or worse, discount it, and you are marking down a product with no stock risk at all, often one of your higher-margin lines.
How SellRank avoids the made-to-order trap
SellRank separates two questions that most stock logic merges. 'Available to buy' answers whether a customer can purchase the product right now, and it includes made-to-order products, so they stay in in-stock slots and are never buried as sold out. 'Inventory, total units' counts physical stock only. Its sell-through and stock-cover logic is written never to mistake made-to-order or untracked items for overstock. The distinction came straight from merchandising a live jewellery catalogue, where made-to-order pieces sit alongside stocked ones in the same collections, and getting it wrong would have pushed the wrong pieces into clearance.
If you use the nominal-quantity approach for made-to-order items, it is still worth tagging them, for example 'made-to-order', so you can exclude the tag explicitly from any clearance slot. Belt and braces costs nothing.
Step 1: Give overstock exposure before you give it a discount
The first intervention for real overstock should be placement, not price. It costs no margin, it is reversible in one click, and it tests the question that matters: is this product unwanted, or unseen?
In practice, that means a slow-mover slot on the full-price collections where the overstocked products belong. Keep it modest and keep it below the first row, which has to convert. A typical shape:
- Slot 1, first 20%: in-stock products ranked by Conversion % or Revenue. The first screen still sells.
- Slot 2, next 10 to 15%: in-stock products ranked by days of stock cover, highest first, excluding made-to-order items. This is the exposure slot.
- Everything else: your normal ranking, with sold-out products parked at the end.
Run it for two to three weeks. Because SellRank keeps re-sorting while the smart sort is enabled, products that start to sell see their cover fall and leave the slot on their own, making room for the next-most-overstocked product. The slot is self-rotating, which a hand-dragged manual order can never be.
Step 2: Read the result before cutting price
After the exposure period, the overstock list splits into three groups. Some products sold noticeably better once they were visible; they were never unwanted, and they can stay in rotation at full price. Some sold a little better; they may need a modest markdown, not a deep one. And some did not move at all despite prime placement; those are the products that genuinely need a discount, and now you have evidence rather than a hunch.
Comparing like with like matters here. Compare each product's rate of sale during the exposure period with its rate before, and look at conversion from view to order rather than raw orders, because the exposure slot itself brings more views. In SellRank the version that introduced the exposure slot is kept alongside the one before it, and the performance view attributes daily views, add-to-carts, orders and revenue to whichever version was live, so the collection-level effect is visible too. Per-version performance is on Grow and above.
Step 3: Discount only what exposure could not move
Now the markdown is targeted. Instead of 30% off the category, you discount the products that proved they need it, and you can size the discount to how badly they need it. The rest of the category keeps its full price, and customers are not trained to wait for a sitewide event.
This is also where margin data earns its place. Two products with identical overstock can carry very different margins. A product with a 70% margin can take a deep cut and still make money; a product with a 25% margin may be better held, bundled or returned to a supplier than discounted to below cost.
Profit-aware clearance ranking on Scale and up
On SellRank's Scale and Max plans, profit-aware ranking adds Margin % and estimated Profit to the strategy list, read from the cost per item on your products. For clearance, that opens up a few useful moves:
- Rank the clearance collection so the products that still carry healthy margin at their sale price lead, so the first screen of your sale makes money rather than just moving units.
- Blend margin with days of cover in a Balanced score, so the slot favours products that are both overstocked and profitable to clear.
- Spot products whose sale price has pushed margin close to zero, and decide deliberately whether they belong in the sale at all.
Profit sorting needs cost per item filled in. Where it is missing, SellRank reports profit and margin as zero rather than guessing, and the label says 'est.' for that reason. On many stores the first job of clearance planning is filling in costs on the overstocked products, which is worth doing anyway. Our earlier post on sorting by profit, not just sales covers the wider case for margin-aware merchandising.
How to order a Shopify clearance collection
Most Shopify clearance collections are sorted by discount, deepest first, or by price, low to high. Both put the least desirable stock on the first screen. A clearance page still has to convert, and it converts on the same things any collection does: products people want, in sizes and stock they can buy.
- Slot 1, first 25%: in-stock clearance products ranked by Conversion %. The first screen should be the sale items shoppers actually buy.
- Slot 2, next 25%: in-stock clearance products ranked by days of stock cover, highest first. This is where the deepest overstock gets strong placement.
- Everything else: Discount, deepest first, which is what sale shoppers who scroll expect to find.
- Park sold-out products at the end, so the page does not fill with items that cleared yesterday.
- On Scale and up, blend Margin % into slot 1 so the first screen is profitable as well as popular.
If the clearance collection is a seasonal event rather than a permanent page, build its order as a draft, schedule its go-live for the start of the sale and set a revert date for the end. The previous order returns automatically. Drafts, scheduled publishes and auto-revert are on Grow and above.
A worked clearance example
Illustrative scenario: an outdoor apparel store ends summer with too many lightweight jackets. There are 22 styles in the 'Jackets' collection. Ranked by days of stock cover, six stand out with more than 200 days of cover and low sell-through. Two of the six are made-to-order custom-printed jackets with nominal stock of 999, which a naive tool would have placed at the top of the overstock list. SellRank's stock logic leaves them out, so the real list is four styles.
Week one to three: the four styles go into a 15% exposure slot on 'Jackets' and 'Outdoor', just below a conversion-ranked first row. Two of them start selling at several times their previous rate once they are visible above the second scroll. Their cover falls and they drop out of the slot on their own.
Week four: the remaining two styles have barely moved despite prime placement. They go into the clearance collection. One carries a 65% margin and takes a 40% markdown comfortably. The other carries a 30% margin, and a 40% cut would put it close to break-even, so it takes a smaller discount and is bundled with a best-selling accessory instead.
The result: two styles cleared at full price, two cleared with targeted discounts sized to their margins, and the rest of the category never discounted at all. None of this required anyone to drag products around by hand.
Protecting your winners with stock cover
Clearance gets the attention, but days of stock cover is just as useful at the other end of the list. Every store has a handful of products that sell faster than they can be replenished. On a sales-ranked collection, those products sit at the top until the moment they sell out, and then they sit at the top sold out. The page's most valuable positions go dead exactly when traffic is highest.
A better pattern is to let cover act as an early warning. When a best-seller's cover falls below your replenishment lead time, it should stop being the first thing every visitor sees. It still sells, from lower on the page and from direct links, but the scarce remaining stock is not burned through by the widest possible audience, and the first row is handed to products that can take the demand. When the restock lands, cover rises and the product climbs back up.
In SellRank you can express this with an in-stock condition on the top slot combined with a ranking that blends demand with cover in a Balanced score, and with out-of-stock push-down as the backstop. Because the order keeps refreshing as stock changes, the adjustment happens without anyone watching the stock report.
When exposure and discount both fail
Some stock will not clear on your own storefront at any sensible price. It is better to recognise this early than to let it drag the clearance collection down for a year. The options, roughly in order of how much value they recover:
- Bundling with a best-seller, so the slow product rides on a product that already converts. This often recovers more than a deep standalone discount.
- Gift with purchase above an order threshold, which lifts average order value and moves units without a visible price cut.
- Wholesale or outlet channels, including marketplaces, where the product reaches a different audience without discounting your main store.
- Returning to the supplier or swapping for other stock, where the terms allow it.
- Donation, which at least recovers storage space and, in some jurisdictions, has tax treatment worth asking your accountant about.
Whatever you choose, get the product out of the prime positions of your main collections. A collection page carrying dead stock near the top is paying twice: once for the stock and again in lost conversion on the products around it.
Using clearance data to buy better next season
The cheapest clearance is the one you never need. Every exposure test you run produces evidence that belongs in your next buying decision: which styles were unwanted, which were merely unseen, which needed a discount and how deep. Keep that list. Styles that sold through at full price once they were visible were under-merchandised, not over-bought, and the fix next season is placement rather than a smaller order. Styles that failed exposure and needed deep cuts are the ones to buy more cautiously.
Because each change in SellRank is a published version with its own performance record, the evidence does not live in somebody's memory. You can go back to the version that introduced last season's exposure slot and see what the collection did while it was live.
Sell-through and cover by type of store
- Apparel and footwear: sell-through is the core metric, and size breaks complicate it. Treat broken styles separately, as covered in our fashion collection sorting guide.
- Beauty and consumables: expiry dates make cover critical. A product with more cover than shelf life is overstock even if it sells steadily.
- Homeware and furniture: high unit values and bulky storage make cover expensive. Exposure testing is especially worthwhile because a single extra sale matters.
- Jewellery and made-to-order goods: the made-to-order trap is everywhere. Separate physical stock from availability before trusting any cover figure.
- Gifts and seasonal goods: cover must be read against the calendar. Ninety days of cover in October is fine; ninety days in late December is a problem.
Clearance merchandising mistakes to avoid
- Putting overstock in the very first row. It costs more in lost conversion than it saves in markdown. Use a slot below the first screen.
- Judging a product as unwanted without checking its position. Low sales at position 60 is evidence of invisibility, not of weak demand.
- Running clearance logic on products that are too new to have a fair rate of sale.
- Letting made-to-order or untracked items into overstock calculations.
- Discounting without looking at margin, so some clearance sales lose money on every unit.
- Leaving the clearance order running after the sale ends. Schedule the revert when you schedule the sale.
- Sorting the clearance page by deepest discount, so the first screen is the stock nobody wanted at any price.
How often to review overstock
The ordering itself should refresh continuously, which is what automation is for. The strategy should be reviewed on a slower cycle. For most stores, a monthly overstock review works: rank each major collection by days of cover, check what has moved in and out of the exposure slots, and decide which products graduate to a markdown. Seasonal categories need an extra review four to six weeks before season end, when there is still time for exposure to work at full price.
Because SellRank keeps every published order as a version, each review can compare the current configuration with the previous one on the performance view rather than from memory.
A monthly overstock review checklist
- Rank each major collection by days of stock cover in the preview and note the top ten products on each.
- Remove made-to-order, pre-order and too-new products from the list before drawing conclusions.
- Check each remaining product's current position. Anything below the second scroll has not had a fair test.
- Move untested overstock into the exposure slot, and leave it there for two to three weeks.
- For products that have already had exposure, compare their rate of sale and view-to-order conversion before and during the test.
- Send the products that failed exposure to targeted markdown, sized to their margin, or to bundles and outlet channels.
- Check winners with low cover against their replenishment lead time and make sure the top slot is not about to break.
- Record the decisions next to the version that implemented them, so next month starts from evidence rather than memory.
Explaining clearance decisions to the rest of the business
Clearance is where merchandising, finance and buying meet, and they often disagree. Finance wants the stock off the balance sheet; buying wants to know what went wrong; marketing wants a sale to promote. Sell-through and days of cover give everyone the same language. 'These four styles have more than 200 days of cover after three weeks of prime placement at full price' is a far easier sentence to agree on than 'these feel slow'.
The per-version record helps here too. When someone asks whether the exposure slot cost the collection anything, the answer is on the performance view: the collection's views, add-to-carts, orders and revenue while the exposure version was live, next to the version before it, over comparable periods. Our post on how to prove a collection order change worked explains how to read that comparison fairly, including how to allow for ad spend and seasonality.
Choosing the right tool for clearance merchandising
Shopify's built-in sorts have no concept of sell-through or stock cover; 'Best selling' actively buries overstock. A hand-maintained manual order can implement everything above, but the exposure slot needs re-arranging every time a product sells through, which is precisely the job nobody has time for. Client-side reordering scripts change the display without changing the real order. Server-side sorting apps vary in whether they offer stock-based strategies at all, whether they understand made-to-order stock, and whether they gate inventory logic behind their most expensive plan.
SellRank puts Sell-through and days of stock cover on every plan, including Free, handles made-to-order stock correctly, composes exposure slots with conditions, adds profit-aware ranking on Scale and Max, and shows version by version whether the change worked. For clearance, we think that combination is the strongest option available to a Shopify merchant, and all of it can be checked on your own catalogue.
- Free: one collection, full editor and preview, sell-through and stock cover, one live sync.
- Grow, $29 a month: 15 collections, unlimited syncs, drafts, scheduled publishes and auto-revert, versions and performance analytics.
- Scale, $69 a month: 75 collections, global rule, GA4 backfill, profit metrics including Margin % and estimated Profit, and drift alerts.
- Max, $249 a month: unlimited collections and priority support.
Find your real overstock this week
Open your most overstocked category in SellRank's preview, rank it by days of stock cover, and look at the top of the list. Check it for made-to-order items, check where each product currently sits on the page, and you will have a clearer clearance plan in ten minutes than a blanket markdown would give you in a month. Then add an exposure slot, sync it, and give it three weeks before you touch a price.
SellRank installs in about five minutes, touches no theme files, and is free on one collection. See every strategy, slot condition and plan on the SellRank product page.
