Your Worst-Performing SKUs Are Not the Ones With the Lowest Revenue
The quarterly SKU review takes about twenty minutes at most brands. Someone pulls a revenue report, sorts it descending, draws a line somewhere in the bottom third, and asks whether anything below the line should be discontinued. It feels rigorous. It is the wrong analysis.
The $420K SKU with a 19% trade rate and recurring chargebacks might be losing money after channel costs. The $180K SKU with clean compliance and steady velocity might be your best performer. Revenue rank gets this backwards — it measures what sold, not what it cost to sell.
Strategy& estimates that 35% of CPG SKUs drive zero incremental profitability, and the bottom 15% actively destroy gross margin. At Nestle, only 11% of 100,000 SKUs account for 80% of sales. The long tail is not just unproductive — it actively costs money in complexity, shelf dilution, and trade spend allocated to products that will never move fast enough to justify it.
What revenue rank misses
Revenue measures the top line. It does not capture the four cost layers that determine whether a SKU actually earns money.
Trade cost per SKU. Trade spend runs 15–30% of gross sales for CPG brands in conventional grocery — but it is not evenly distributed across the catalog. A SKU placed in three high-trade-rate retail accounts carries a different cost structure than one sold primarily through distribution. The blended portfolio rate hides the SKUs that are individually underwater.
Chargeback load. Compliance chargebacks — labeling errors, dimension mismatches, ASN failures — cluster around specific products, not around the catalog at random. A SKU with a case-cube error in the product master generates a chargeback on every shipment to that retailer. The chargebacks are coded to the shipment, not the product, so the per-SKU cost is invisible in the standard report.
Velocity trajectory. Current units per store per week matters, but the direction matters more. A SKU at 2.8 UPSPW and declining is a different proposition than one at 2.0 and climbing. Revenue rank is a snapshot. Velocity trajectory is the trend — and the trend is what the buyer is watching at the next line review. A catalog full of declining velocities gives the buyer a reason to question your entire set, not just the weak performers.
Channel concentration. A SKU generating 85% of its revenue through a single retailer is one buyer decision away from a write-off. The portfolio review that ignores concentration risk is pricing in certainty that does not exist.
Why this matters more than one review cycle
The wrong SKUs in the catalog do not sit quietly. They consume warehouse space, label inventory, broker attention, buyer goodwill, and trade dollars that would produce better returns directed at products that move.
And there is a cost that does not appear on any internal report: what the buyer sees. A bloated catalog with weak velocities across half the line tells the category manager you are not managing your portfolio. At the next line review, that buyer is less likely to give you the additional slot you asked for — not because the new product is bad, but because the existing set suggests you will not support it.
L.E.K. Consulting's research suggests that SKU rationalization typically delivers a 65-to-90-basis-point margin increase. At $25M in revenue, that is $163K–$225K recovered annually — not from selling more, but from carrying less. Fewer SKUs means less complexity in production, warehousing, and retailer compliance, and more trade dollars concentrated on products that earn their shelf space.
PepsiCo recently announced cutting 20% of its SKUs to protect what it calls "hero SKUs that drive trips and basket add-ons." The articulation is correct: protect what drives traffic, cut what does not. The method for identifying which is which — contribution per SKU after trade cost, chargeback load, and velocity — is what most brands have not built. SKU rationalization is one of the Ten Decisions — each quantifiable, all answerable from data the brand already has.
What the audit looks like
The portfolio audit scores every SKU across the four dimensions revenue rank ignores.
Score by contribution, not revenue. Pull trade spend by SKU from your trade planner — whether that is a dedicated tool or the spreadsheet your broker maintains. Pull chargebacks by SKU from the deduction ledger. Calculate contribution margin after channel costs — not gross margin, which excludes the cost of earning that margin through a specific channel.
Add the trajectory. A SKU's trailing 26-week velocity slope separates products losing their audience from products building one. The current number tells you where you are. The slope tells you where you are going.
Flag concentration risk. Any SKU where more than 70% of volume flows through a single account is one delisting away from a write-off. The audit surfaces this so the rationalization decision reflects the full risk picture, not just the current economics.
The output is a quadrant map — kill, fix-or-kill, watch, and grow — with the economics behind each classification. The fix-or-kill category is where the highest-return work sits: products where the consumer appeal is real but the economics are broken by a specific, fixable factor — a trade rate negotiated too high, a channel placement that generates chargebacks, or a format that creates recurring compliance cost. Fix the factor and the math changes.
See it worked through
I built a synthetic $25M specialty food brand — Cinderhaven Provisions, 50 SKUs across ten sell-through channels — to run this audit end to end. Real data shapes, real cost structures, no client data exposed. The result: 19 of 50 SKUs were kill candidates and 22 were fix-or-kill when scored by contribution after trade cost, chargeback load, and velocity trajectory. A standard sort-by-revenue review would have flagged a different set.