You grew revenue 35% last year. You also grew your SKU count 60%. And somehow, despite record top-line numbers, your margins are thinner, your warehouse is slower, and your demand planner just quit.

That’s not a coincidence. That’s SKU proliferation doing exactly what it does.

The Problem Nobody Budgets For

Every new SKU feels like a growth move. A new flavor. A retailer-exclusive size. A limited-edition collaboration. Each one has a business case, a revenue projection, maybe even a purchase order from a buyer who asked for it.

What none of them have is a line item for the operational drag they create the moment they enter your catalog.

Here’s what actually happens when you add SKU #301 to a catalog that had 300:

• Your warehouse now has one more pick location to maintain, one more bin to count, one more slot competing for prime real estate near the pack station
• Your demand planner has one more time series to forecast — with zero historical data and no analog to train from
• Your purchasing team has one more MOQ to negotiate, one more lead time to track, one more vendor relationship to manage
• Your marketing team has one more product to photograph, write copy for, and position across every channel
• Your customer service team has one more product to learn, one more return reason to handle, one more FAQ to maintain
• Your finance team has one more line item to track inventory value, landed cost, and margin on

None of these costs show up on the SKU’s P&L. All of them show up on yours. And they compound. SKU #302 doesn’t just add its own overhead — it makes #1 through #301 slightly harder to manage, too. Forecasting models lose accuracy as they get spread thinner. Warehouse slotting gets less optimal. Marketing budgets get diluted across more products competing for the same customer’s attention.

What a SKU Actually Costs You

Most brands calculate SKU profitability like this: revenue minus COGS minus maybe shipping. That math is wrong by 30–60%.

The true cost of carrying a SKU includes at least seven categories that most operators ignore entirely:

Cost Category	What It Includes	Typical Range (Annual)
Warehousing	Bin space, slotting, re-slotting, cube utilization loss	$800–$3,200 per SKU
Inventory carrying	Capital cost, insurance, shrinkage, obsolescence risk	20–30% of avg. inventory value
Forecasting complexity	Analyst time, error rate increase, safety stock buffers	$400–$1,500 per SKU
Purchasing overhead	PO management, vendor comms, MOQ compliance	$300–$900 per SKU
Marketing dilution	Content creation, catalog space, ad spend fragmentation	$500–$4,000 per SKU
Customer service	Training, returns processing, complaint handling	$200–$800 per SKU
Systems & data	Product setup, sync across channels, catalog maintenance	$150–$600 per SKU

Add those up for a mid-market CPG brand running 400 SKUs, and the catalog overhead — the cost of simply having all those SKUs exist — runs $1.2M to $4.4M per year. That’s before you sell a single unit.

The warehousing line alone surprises most operators. A single SKU in a multi-level racking system doesn’t just occupy one bin. It needs a forward pick location and a reserve location. When velocity drops, that forward pick location becomes dead space — premium floor area generating almost no throughput. At $8–$15 per square foot per month in a modern 3PL, a dormant SKU occupying 4 square feet of forward pick costs $384–$720/year in rent alone, not counting the labor to count it, the system overhead to track it, or the opportunity cost of that space holding a faster-moving product.

The Proliferation Lifecycle

SKU proliferation doesn’t happen all at once. It follows a predictable pattern that shows up in almost every brand that scales past $10M:

Stage 1, the Core — 0 to 50 SKUs. Every product earns its place. The team knows every SKU by name. Forecasting is simple because you can hold the whole catalog in your head. Warehouse layout is clean. Life is good.

Stage 2, the Expansion — 50 to 150 SKUs. New sizes, new flavors, new bundles. Each addition has a clear revenue justification. Operational complexity grows, but it’s manageable because the team is still small enough to compensate with hustle and institutional knowledge.

Stage 3, the Accommodation — 150 to 300 SKUs. This is where it turns. Retailer-specific exclusives start appearing. Channel-specific bundles that exist because a buyer asked for them. Seasonal variants that launched for Q4 and never got sunsetted because nobody wanted to be the one to kill them. The catalog starts serving other people’s strategies instead of yours. Forecasting accuracy drops below 70%. Warehouse pick times start creeping up. Nobody has a clear picture of which SKUs actually make money after fully loaded costs.

Stage 4, the Drag — 300+ SKUs. Half the catalog generates less than 5% of revenue. But nobody wants to kill anything because “what if that buyer reorders?” or “we just redesigned the packaging” or “it’s only costing us a little bit.” The long tail is now a long anchor, and it’s dragging down the metrics of your entire operation.

If you’re recognizing stage 3 or 4, you’re not alone. A 2024 analysis by IHL Group found that SKU complexity costs mid-market retailers and brands an average of 3.2% of revenue annually — not from the products themselves, but from the operational overhead of managing them.

Calculating SKU-Level True Profitability

Before you can rationalize your catalog, you need to know what each SKU actually contributes. Here’s the formula that accounts for fully loaded costs:

SKU True Contribution = Revenue
                       - COGS
                       - (Allocated Warehousing Cost)
                       - (Avg Inventory Value × Carrying Cost Rate)
                       - (Allocated Forecast Complexity Cost)
                       - (Allocated Purchasing Overhead)
                       - (Allocated Marketing Spend)
                       - Channel Fees & Commissions
                       - Freight (Inbound + Outbound)

Example — "Organic Lavender Hand Soap, 12oz":

  Revenue:                    $84,000
  COGS:                      -$33,600  (40%)
  Warehousing allocation:     -$1,800
  Carrying cost (25% × $14K): -$3,500
  Forecast complexity:          -$900
  Purchasing overhead:          -$500
  Marketing allocation:       -$2,200
  Channel fees (15%):        -$12,600
  Freight:                    -$8,400

  Gross margin (simple):      $37,800  (45.0%)
  True contribution:          $20,500  (24.4%)

That 20-point gap between perceived margin and actual contribution is where SKU proliferation hides. Multiply it across 50 underperforming SKUs and you’re looking at hundreds of thousands in phantom margin — revenue that looks profitable in your accounting system but is actually being consumed by the operational overhead of keeping those products alive.

The allocation methodology matters, too. Some costs (warehousing, carrying) can be directly attributed by SKU. Others (forecasting complexity, marketing dilution) need to be allocated. The simplest defensible approach: divide total department cost by active SKU count, then weight by the amount of attention each SKU actually consumes. A SKU that requires a custom forecast model and a dedicated marketing campaign gets a higher allocation than one that rides in the same product family.

The Rationalization Framework

Killing SKUs is politically harder than launching them. Every SKU has a champion — the sales rep who promised it to a buyer, the founder who designed the packaging, the marketing manager who just shot the photography. You need a framework that depersonalizes the decision and replaces opinion with data.

The Four-Quadrant Sort

Plot every SKU on two axes: true contribution margin (vertical) and velocity in units sold per month (horizontal). This gives you four buckets:

	High Velocity	Low Velocity
High Margin	Stars — protect and invest	Sleepers — investigate why velocity is low
Low Margin	Workhorses — optimize costs or raise prices	Deadweight — sunset candidates

The Deadweight quadrant is obvious. Low margin, low velocity — these SKUs cost you more to maintain than they contribute. Start here.

But the real wins often come from the Workhorses quadrant. These are SKUs moving volume but destroying margin. They feel important because they generate revenue, but they’re actually subsidized by your Stars. The fix for a Workhorse isn’t always a sunset — sometimes it’s a price increase, a packaging simplification, or a vendor renegotiation. But if you can’t move a Workhorse into positive true contribution within two quarters, it becomes a sunset candidate too.

Sleepers are worth a closer look before you act. A SKU with high margin but low velocity might be a new product that hasn’t found its audience, a seasonal item being measured in the wrong window, or a product with distribution gaps. Give Sleepers a 90-day focused push — better placement, targeted marketing, a promotional test — before classifying them.

The Sunset Criteria

A SKU becomes a sunset candidate when it meets two or more of these conditions:

1. True contribution margin is below your catalog median
2. Monthly velocity has declined for three or more consecutive quarters
3. It has fewer than two active sales channels
4. It requires a dedicated MOQ that ties up more than $15K in working capital
5. Customer return or complaint rate exceeds 8%
6. It was created to satisfy a single buyer or channel that no longer orders at volume
7. It cannibalizes more than 40% of a higher-margin SKU’s sales in the same category

The Sunset Process

Don’t just discontinue and hope. A structured sunset protects revenue, preserves buyer relationships, and captures whatever residual value remains in the inventory.

Weeks 1–2, financial validation. Run the true contribution calculation. Confirm the SKU qualifies. Model the revenue impact — if the SKU does $50K/year but costs $35K to carry, the net loss from cutting it is only $15K, not $50K. Most operators dramatically overestimate the cost of cutting a SKU because they look at top-line revenue rather than true contribution.

Weeks 3–4, stakeholder notification. Tell your sales team, key buyers, and customer service. For retail partners, offer a substitution SKU from the same product family. For DTC customers, send a “last chance to buy” email — which often generates a meaningful bump in sell-through that helps clear remaining inventory. Frame the communication around “making room for what’s next” rather than “this product failed.”

Weeks 5–8, inventory depletion. Stop reordering. Run the SKU down to zero through normal sales, promotional pricing, or bundle inclusion with higher-velocity products. Do not destroy inventory if you can liquidate it at any positive margin. If you have more than 90 days of supply remaining, consider offering it to a closeout channel — you’ll get 20–40 cents on the dollar, but you’ll free up warehouse space and working capital months earlier.

Weeks 9–10, system cleanup. Deactivate across all channels simultaneously. Archive the listing — don’t delete it, because you’ll want historical data for future forecasting. Update your product feed, catalogs, price lists, and marketing materials. Remove from warehouse slotting and release the bin location for reallocation.

Week 11 and beyond, measurement. Track what happens to the metrics that matter: warehouse pick rate, forecast accuracy, purchasing team bandwidth, working capital availability. You’ll almost always see improvement within one quarter, and often within the first month. The absence of noise makes everything else run cleaner.

When Killing a SKU Is the Wrong Call

Not every low-performer should die. Some SKUs serve a strategic function beyond their own P&L, and cutting them would cost you more than keeping them:

Gateway SKUs introduce customers who then buy higher-margin products. Measure this by tracking first-purchase-to-repeat-purchase paths in your order data. If 40% of your best customers entered through a specific SKU, that SKU’s value isn’t in its own contribution margin — it’s in the customer lifetime value it generates downstream. The gateway test: remove the SKU from the catalog, and do new customer acquisition economics get worse?

Assortment anchors are products that a key retail partner requires to maintain your shelf set or planogram placement. Losing one SKU might cost you three facings. Before cutting anything in a retail assortment, confirm with your buyer what the minimum SKU count is to hold your shelf position. Sometimes you can substitute rather than eliminate.

Seasonal tentpoles drive 60%+ of their annual volume in a six-week window. Their annualized metrics look terrible — low average velocity, high carrying cost relative to sales — but their in-season contribution may be excellent. Evaluate these on peak-period economics, not annual averages. If a holiday SKU generates $40K in contribution during November and December but costs $6K to carry through the other ten months, it’s still a $34K winner.

The Pushback You’ll Get (and How to Handle It)

Every rationalization project runs into the same objections. Having answers ready keeps the process from stalling.

“We’ll lose revenue.” Yes — 4–6% of top-line, typically. But you’ll gain 5–8 points of margin. Run the true contribution numbers for the sunset candidates and show the team what “revenue” actually means after loaded costs. Most of these SKUs aren’t generating profit. They’re generating activity.

“Our buyer specifically asked for this SKU.” Confirm that the buyer is still ordering it. Pull the last 12 months of POs from that account. Half the time, the buyer who requested the product has moved on, and the account hasn’t reordered in two or three quarters. For active relationships, offer a substitution and let the buyer weigh in before you cut.

“What if a customer wants it?” Look at the actual purchase data. If a SKU has fewer than 50 unique purchasers in the trailing 12 months, the impact of sunsetting it is measurable and manageable. You can notify those customers directly, offer alternatives, and in many cases convert them to a similar product with better margins.

“We just invested in new packaging for this.” Sunk cost. The packaging investment doesn’t change whether the SKU generates positive true contribution going forward. If anything, a recent packaging refresh on a low-performer is a warning sign — you’re investing in a product the market has already voted on.

The Velocity-Complexity Ratio

Once you’ve done the initial rationalization, you need an ongoing metric to prevent proliferation from creeping back. New product launches are exciting. Sunsets are not. Without a guardrail, your catalog will re-bloat within 18 months.

The metric that works best is simple:

Velocity-Complexity Ratio (VCR) = Total Units Sold / Active SKU Count

Example:
  Brand A: 1,200,000 units / 180 SKUs = 6,667 units per SKU
  Brand B: 1,200,000 units / 420 SKUs = 2,857 units per SKU

  Brand A and Brand B have identical revenue.
  Brand B's operational cost per unit is 35–50% higher.

  If Brand B rationalized to 250 SKUs:
  VCR improves to 4,800 — a 68% increase in catalog efficiency.
  Estimated annual overhead savings: $380K–$680K.

Track your VCR monthly. When it drops for two consecutive months without a corresponding revenue increase, you’re adding complexity faster than you’re adding value. That’s the signal to pause launches and run a rationalization cycle.

Healthy benchmarks by category:

Category	Strong VCR	Warning Zone	Danger Zone
Food & Beverage	> 8,000	4,000–8,000	< 4,000
Beauty & Personal Care	> 5,000	2,500–5,000	< 2,500
Home & Kitchen	> 3,000	1,500–3,000	< 1,500
Supplements & Wellness	> 6,000	3,000–6,000	< 3,000

The Launch Gate

The best way to manage SKU proliferation is to prevent it at the source. Before any new SKU enters your catalog, it should pass through a launch gate — a standardized review that forces the business case to include operational costs, not just revenue projections.

The gate requires answers to five questions:

1. What is the projected true contribution margin after fully loaded costs? It must exceed your catalog median to justify the added complexity.
2. Which existing SKU will this cannibalize, and by how much? If it cannibalizes more than 30% of an existing SKU’s volume, you should sunset the old one simultaneously.
3. Does this require a new vendor, new packaging format, or new warehouse configuration? Each “yes” adds $2K–$8K in setup costs that must be recovered in year one.
4. What is the kill criteria? Define the velocity and margin thresholds that trigger an automatic sunset review at 6 and 12 months post-launch. Write them down before you launch, when you’re still objective.
5. Who owns the sunset decision if it underperforms? Name a person, not a committee. Not “the leadership team.” Not “we’ll revisit.” A name, a timeline, and a threshold.

That fifth question matters most. SKUs without a named owner never get killed. They sit in your catalog, quietly consuming warehouse space, forecast cycles, and working capital until someone finally does the analysis you’re reading right now — usually two years too late.

One practical tip: build the launch gate into your existing product development workflow, not as a separate review. If new SKUs already go through a design review or a marketing brief, add the five questions to that existing checkpoint. The goal is to make operational cost analysis a natural part of the launch conversation, not a bureaucratic hurdle that people route around.

A Real Catalog Cleanup, By the Numbers

A $28M home goods brand we worked with had 340 active SKUs when they started this process. They’d grown from 80 SKUs three years earlier, mostly through retail expansion and line extensions. Revenue had nearly doubled, but operating margin had actually declined by two points — classic stage 4 proliferation.

Over one quarter, they ran every SKU through the four-quadrant sort and the true contribution calculation. Here’s what they found:

• 112 SKUs (33%) were Deadweight — low margin, low velocity, no strategic rationale
• 47 SKUs (14%) were Workhorses — high volume but negative true contribution after fully loaded costs
• The bottom 150 SKUs generated just 8% of revenue but consumed 41% of warehouse bin space
• 23 SKUs had zero units sold in the trailing 90 days — pure carrying cost with no offsetting revenue
• 31 SKUs existed solely because of a single retail buyer relationship that had gone dormant

They sunset 89 SKUs over four months using the structured process. The results after two full quarters:

• Revenue declined 4.2% (from losing the long-tail SKUs)
• Gross margin improved 6.8 points (from eliminating negative-contribution products and reducing overhead allocation across fewer SKUs)
• Warehouse pick accuracy went from 96.1% to 99.2%
• Demand forecast accuracy (MAPE) improved from 38% to 24%
• Average order processing time dropped 22%
• Working capital freed up: $1.1M in inventory that no longer needed to be carried

The net impact was a $1.6M improvement in operating profit on 4% less revenue. That’s the math that makes SKU rationalization one of the highest-ROI operational projects a scaling brand can run — and it’s the reason the best operators treat their catalog like a portfolio, not a collection.

Where to Start Monday Morning

Pull your full SKU list from your OMS or ERP. Sort by trailing-90-day units sold. Look at the bottom 20%.

For each SKU in that bottom quintile, answer two questions: what does it cost us to keep this alive, and what would we actually lose if it disappeared tomorrow? If the answer to the first question is bigger than the answer to the second — and it usually is — you’ve found your first sunset candidates.

If you can’t answer those questions with real numbers, that’s the first problem to solve. Your ERP and OMS have the sales data. Your 3PL or WMS has the warehousing costs. Your marketing platform has the spend allocation. The challenge is connecting them in a way that produces true contribution by SKU, not just gross margin.

CommerceOS was built to surface exactly this kind of cross-system visibility — pulling order, inventory, warehouse, and channel data into a single view so you can see true SKU economics without maintaining a 47-tab spreadsheet. If you’re running the analysis manually and want to see what it looks like automated, book a demo and bring your SKU list. We’ll run the first pass together.