Let’s talk about how we can build your commerce project — tailored to your business, powered by Mercur
SKU rationalization is the process of deciding which stock keeping units earn their place in your catalog and which ones are costing you more than they return. The SKU rationalization process runs on data analysis and ends in a portfolio decision that reshapes your inventory management for the year ahead.
Most commerce teams grow the same way for years. More SKUs, more variants, more pack sizes, more channel-specific listings. Each addition looks like upside because it can only add revenue, never subtract it.
The arithmetic stops working at some point. Every SKU you add draws on the same forecasting attention, working capital, operations team, and the marginal one draws more than it gives back.
In this article, you will learn:
- What is SKU rationalization, and how does it differ from product rationalization?
- Why does adding more SKUs stop increasing revenue?
- How do you run the process in 4 steps?
- What goes wrong, and how do you get past it?
- Can you keep the range without carrying the inventory?
Key takeaways
- SKU rationalization decides which products stay, which go, and which move to someone else's balance sheet.
- Adding SKUs raises revenue in the early stages and suppresses it later, because complexity cost grows faster than the sales each new SKU brings in.
- The cost lands in places nobody attributes to the catalog: forecasting error, working capital, warehouse handling, content production, and support load.
- A defensible cut needs four inputs: SKU performance, demand signals, product mix coverage, and lifecycle stage.
- Cutting is one option. Moving the long tail to third-party sellers keeps the range on your site without keeping it on your books.
What is SKU rationalization?
SKU rationalization is a structured review of your catalog that ends in a decision for every SKU: keep, watch, discontinue, or source differently. The inputs are sales performance, margin, demand signals, and where each product sits in its lifecycle.
The exercise is often mistaken for cost cutting. It is closer to portfolio management, because the goal is a catalog where each product carries its own weight rather than a shorter catalog for its own sake.
The benefits of SKU rationalization land in three places: working capital released, operational efficiency recovered, and overall profitability improved on the products you keep.
Deloitte's 2026 Global Consumer Products Industry Outlook found that about half of the organizations intend to rationalize SKUs to reduce complexity and stay closer to changing consumer needs. The practice has moved from periodic housekeeping to a standing item on the operating agenda.
A comprehensive understanding of your own catalog is what separates informed decisions from a round of guesswork with a spreadsheet attached.
SKU vs UPC: what each code identifies
A SKU is your own internal identifier. You create it, you control its format, and it describes the product at whatever level of detail your operation needs: size, colour, pack, location.
A UPC is a global identifier issued through GS1 and used by everyone who handles the product. Two retailers selling the same item share its UPC and will almost never share its SKU.
The distinction matters for SKU analysis. Your SKU data tells you how a product performs in your business, while UPC data lets you compare that performance against the same product elsewhere.
What is product rationalization, and how does it differ?
Product rationalization asks whether a product or a whole product line belongs in the portfolio at all, taking in manufacturing process, production costs, and strategic fit.
SKU rationalization works inside that answer. A product can be worth keeping while three of its five pack sizes are not.
In practice, the two run together. You decide which product lines stay, then decide how many SKUs each line needs to serve its demand.
What does SKU rationalization look like in practice?
A hardware retailer carries the same fastener in six pack sizes because three buyers once asked for an unusual quantity. Five of the six sell predictably. The sixth turns over twice a year and has been reordered automatically for four years.
A fashion brand keeps a colourway that sells only in one market, in one size run, at full price never. It stays in the catalog because discontinuing it requires a decision and reordering it does not.
Both cases share the same mechanism: the SKU was added by a decision and is kept by the absence of one. SKU rationalization is what turns that absence back into a decision.
Why does adding more SKUs stop increasing revenue?
Adding a SKU has an obvious upside and a hidden cost, and the two scale differently.
The upside is linear, because a new SKU sells what it sells. The cost compounds, because every SKU draws on shared capacity that does not grow when the catalog does.
Past a certain catalog size, the marginal SKU consumes more forecasting attention, working capital, and operational handling than the revenue it contributes. Revenue keeps rising for a while, so the effect is invisible in the top line and shows up in margin instead.
The complexity cost nobody budgets for
The cost of a SKU runs well past its cost of goods. It takes a share of everything else:
- a forecast someone has to produce,
- stock someone has to fund,
- a location someone has to pick from,
- content someone has to write,
- support tickets someone has to answer.
The same SKU also spreads through the supply chain, where every extra line adds a forecast, a purchase order, and a receiving event. None of that is coded to the SKU in your P&L, which is why the catalog looks free to extend and why profitability erodes without an obvious cause. A single SKU is cheap, and a thousand marginal SKUs are not, and no report shows you the crossover.
The margin these costs come out of is thinner than most people assume. FMI reports that grocery stores carried an average of 33,248 items in 2025, on an average net profit of 2.1%. An assortment that wide, funded on a two-point margin, leaves very little room for products that break even at best.
Where the diminishing returns start
There is no universal SKU count where returns turn negative. It depends on various factors: purchase frequency, margin, lead times, and how much of your assortment shares components.
What is measurable is the effect of reversing the trend. Bain found that reducing SKU complexity can increase sales growth by 2 to 5 percentage points and margins by 100 to 400 basis points in its 2025 consumer products research. If removing complexity raises growth, the complexity was suppressing growth while it was there.
This is the catalog-side version of the scaling wall described in our comparison of marketplace vs eCommerce growth models. Assortment funded from your own balance sheet turns range into a capital decision, and capital decisions have a ceiling.
How to implement SKU rationalization in 4 steps
The process below produces a defensible decision for every SKU. Run all four steps before cutting anything, because each one catches products the previous step would have removed for the wrong reason.
1. Run the SKU analysis
Start with SKU performance over a period long enough to cover the product's normal buying cycle. Pull unit sales, revenue, gross margin, inventory data, and turnover per SKU, so product performance is visible on every measure at once.
Rank on contribution rather than revenue. A high-revenue SKU on a thin margin and slow turns can contribute less than a small one that sells steadily at full price.
For example, certain SKUs post strong revenue on a margin that barely covers handling, and those are the underperforming products a revenue-ranked report will always rate too highly.
Two groups fall out of this immediately. Low-performing SKUs contribute almost nothing on any measure, and redundant SKUs duplicate another SKU closely enough that buyers treat them as interchangeable.
2. Read demand signals beyond sales history
Historical sales data tells you what sold when it was available and merchandised. It cannot tell you what would have sold, which is how discontinuation decisions go wrong.
Add signals that sit ahead of the sale: search terms with no result, add-to-cart without purchase, availability during the measured period, and where the SKU sat in category navigation. A SKU that was out of stock for half the period has been untested rather than proven weak.
Customer demand and market demand also move at different speeds, so a SKU that looks finished on last year's numbers can still have live demand in a channel you under-serve.
Segment by sales channel while you are there. Customer preferences differ enough between channels that a SKU can be marginal overall and central to one target audience.
3. Evaluate product lines and the product mix
Move up from the individual SKU to the product portfolio. The question at this level is coverage: does the product assortment still answer the range of needs the category is bought for?
Cutting purely on performance produces gaps. Remove every slow entry-price item, and you lose the products that bring first-time buyers into the line. Some SKUs earn their place by what they attract rather than by what they sell.
Map the coverage your existing products already give you before removing any of them. Particular products hold a range together at the entry price or the top of the line, and their own sales line understates what they do.
Check the opposite failure too. A diverse product mix on paper often turns out to be several near-identical products that create customer confusion rather than choice.
4. Place every SKU in its lifecycle
A weak SKU at launch and a weak SKU in decline need opposite decisions. Product lifecycle management supplies the difference by placing each product on the curve from introduction through growth, maturity, and decline.
Read performance against lifecycle stage rather than against the catalog average. A product in decline that still turns predictably can be worth keeping until its replacement is ready, while a launch that has missed twice has told you what you need to know.
Eliminating SKUs in late decline is the least contested part of the programme, because the data has already made the case. Products in late decline are where excess stock accumulates.
Our breakdown of how excess inventory builds up and what it costs covers what happens when that decision is deferred.
The decision every SKU lands in
Four outcomes cover the whole catalog. Every SKU should end the analysis in exactly one of them.
6 common challenges in SKU rationalization, and how to overcome them
Most programmes fail in predictable ways rather than novel ones. Scan the signal column and find the one you recognize.
1. Your data is not decision-grade
Ranking SKUs requires margin per SKU, and margin per SKU requires costs that are attributed correctly. Freight, returns, and promotional spend are the three that usually sit at category level instead.
Fix attribution before ranking. A cut list built on wrong margins will remove profitable products and keep unprofitable ones, and it will do so with great confidence.
2. Sales history does not tell you why a SKU sells
Two SKUs with identical sales can have opposite futures. One sells because buyers want it and one sells because it sits in the default position on the category page.
Separate the two before deciding. Merchandising position, availability, and promotional history explain more of the variance than most teams expect.
3. Cutting the tail does not cut the complexity underneath it
This is the failure that surprises people most. Removing listings does not remove the components, specifications, suppliers, and processes that those listings sat on top of.
A catalog can lose a third of its SKUs while keeping every supplier relationship, every packaging format, and every production changeover. Complexity lives below the SKU line, so a cut measured in SKUs can deliver almost none of the saving that was modelled.
Target the layer underneath. Consolidating specifications, pack formats, and suppliers moves cost even when the listing count barely changes.
4. Sales dip and the programme gets reversed
Assortment reduction usually costs something in the short term. Buyers who came for a discontinued item do not always substitute immediately, and the category can show a real decline before it recovers.
Set the expectation before the cut, with a number and a timeframe attached. A programme that is judged on its first month will be reversed in its second, regardless of whether it was correct.
5. Proliferation comes back within a year
A cut is a one-time event, and SKU creation is a continuous process. Remove 500 SKUs without changing how new ones are approved and the catalog returns to its previous size on its own.
Put a gate on the intake. Every new SKU proposal should carry a demand case and a review date, so the next rationalization starts from a smaller backlog.
6. Nobody owns the decision
Analysis is easy to commission and hard to act on, because discontinuing a product creates an identifiable loser inside the business. Without an owner holding the decision, the default is to keep everything and revisit later.
Name the owner before the analysis starts. The deliverable of a rationalization programme is a set of removed SKUs, and a report naming candidates falls short of that.
Best practices for inventory optimization after SKU reduction
The cut is the start. What determines whether the saving survives is how the smaller catalog is run afterwards.
Keep inventory and overhead costs visible
Attribute inventory costs, storage, handling, and support costs to SKUs and product lines on an ongoing basis. A cost that is only visible during a rationalization project will drift back out of view once the project closes.
The instinct after a cut is to reduce inventory costs and stop there. Reduce overhead expenses in the same pass, or the saving stays on paper: store inventory falls while the space, systems, and headcount that carried it stay funded.
Overhead expenses and fixed costs are the ones most likely to survive a SKU cut untouched. Warehouse space, systems, and headcount step down in blocks rather than in proportion, so releasing them takes a separate decision.
A programme set up to cut costs without naming which costs step down tends to deliver a shorter catalog at the same operating expense. Name them, and you reduce costs that were previously treated as fixed.
Watch cash flow rather than only cost. Reduced inventory converts to cash on a delay, and the delay is what determines whether the programme is felt in the business.
Protect availability while you cut
A smaller catalog concentrates demand into fewer SKUs, which raises the cost of a stockout on each one. Lost sales on a core SKU after rationalization undo the margin the cut released.
Raise cover on the products absorbing the substituted demand before the discontinued SKUs sell through. Our breakdown of the seven types of inventory risk covers how to size that exposure.
Product availability on the remaining range decides whether a cut reads as simplification or as absence, and improved customer satisfaction follows the first reading rather than the second.
Build a focused product catalog
A focused product catalog does work that a wide one cannot. High performing products get the merchandising attention, the content, and the shelf position that were previously spread across items nobody was watching.
The gain shows up in marketing efforts as well as operations. Fewer SKUs means a clearer brand identity and a shorter path from category page to purchase, which tends to lift conversion on the products you kept.
Treat SKU optimization as a standing process rather than a project. Overall efficiency improves when the review runs against the same measures every quarter, and better performance on a shorter list compounds across a year.
Set ongoing monitoring against the same measures you used to cut. A rationalized catalog stays rationalized only if the review runs on a schedule rather than on a trigger.
How to cut inventory without cutting the assortment?
Everything above treats the range and the inventory as one decision. They are two, and separating them changes what is possible.
The reason a marginal SKU costs you money is that you own it. You forecast it, fund it, store it, and absorb the loss when it does not sell. Remove the ownership, and most of the cost of a long-tail SKU goes with it, while the listing stays on your site.
What changes when someone else holds the long tail
In a marketplace model, third-party sellers list and fulfill their own inventory. The SKU appears in your catalog, the demand is captured on your domain, and the working capital behind it sits on someone else's balance sheet.
That changes the maths on every SKU in the “source differently” column. A product with real but thin demand stops being a candidate for discontinuation and becomes a candidate for a seller who already stocks it.
The exposure changes rather than disappearing. You take on availability accuracy, fulfillment performance, and catalog data quality instead of forecasting and stock risk.
SKUs with no demand should go, and SKUs with demand you cannot afford to own are the ones worth relisting through a seller.
How Mercur lets sellers carry the long tail
The split above only works if your platform records the stock as belonging to the seller rather than to you. A system that treats all inventory as one pool gives you the operational load of a marketplace without releasing any capital.
Mercur is an open-source marketplace platform under an MIT license, built on the modern Medusa commerce framework. It runs alongside an existing commerce stack instead of replacing it.
Ownership is built into the data model. Every inventory item and stock location is linked to a specific seller, so the long tail sits in the seller's records and the seller's warehouse.
Sellers publish offers against a shared master variant, each with their own SKU, price, and stock. One product in your catalog can carry several sellers' offers, which is what lets a thin-demand SKU stay available without anyone holding deep stock of it.
A single customer cart splits into one order per seller, each with its own shipping method, with commission calculated per line. The code is available on GitHub if you want to see how the seller scoping works, or explore a demo.
Summary: what your SKU count is really telling you
A growing SKU count is usually read as a growing business. It is more reliably a record of decisions that were easy to make and decisions that were never revisited.
SKU rationalization turns that record back into a set of choices. Four inputs decide each one: contribution, demand signals, coverage in the mix, and lifecycle stage.
The part most teams miss is that removal is not the only outcome available. Demand you cannot profitably serve from your own inventory is not automatically demand you have to give up.
If your catalog has grown faster than your margin, and every review ends in a report rather than a decision, the constraint is the business model rather than the discipline.
If you want to work through which part of your tail should be cut and which part should move to third-party supply, talk to a marketplace expert.
FAQ on SKU rationalization
What does it mean to rationalize a SKU?
To rationalize a SKU means to decide, on evidence, whether it stays in the catalog, gets discontinued, or gets sourced a different way. The evidence is its contribution to margin, the demand behind it, the role it plays in the product mix, and its stage in the product lifecycle. The outcome is a decision on that specific SKU rather than a general conclusion about the category.
What is a SKU strategy?
A SKU strategy is the set of rules that decides how SKUs enter your catalog, how they are reviewed, and what removes them. It covers the demand case a new SKU has to make, the measures each SKU is judged on, the review cadence, and who has authority to discontinue. Without those rules, rationalization becomes a periodic clean-up and the catalog returns to its previous size between rounds.
What is SKU reduction?
SKU reduction is the outcome of cutting the number of stock keeping units, while SKU rationalization is the process that decides which ones go. The difference matters because reduction can be achieved by removing anything, including products that were profitable. Reduction is a number, and rationalization is the reasoning that produced it.
How many SKUs should a business carry?
There is no benchmark SKU count, because the right number depends on purchase frequency, gross margin, lead times, and how much your SKUs share components. A business with weekly repeat purchase and high margin can carry a long tail that would sink a business with annual purchase cycles and thin margin. The useful question is whether the marginal SKU still contributes more than it consumes.
Does cutting SKUs reduce revenue?
Cutting SKUs usually costs some revenue in the short term and can raise it over a longer horizon. Buyers of a discontinued item do not all substitute immediately, so a category can decline before it recovers. Agreeing the expected dip and its duration before the cut is what keeps a correct programme from being reversed in its first month.
