Most DTC brands with catalogs over 1,000 SKUs are carrying significant dead weight. Not in the inventory management sense — but in the pricing management sense. A large share of SKUs in a typical e-commerce catalog consume analyst attention disproportionate to the margin they generate or the competitive sensitivity they carry. The Pareto principle applies with unusual sharpness here: in catalogs we've looked at, roughly 15-25% of SKUs account for 80-90% of gross margin contribution. The other 75-85% of the catalog exists, sells occasionally, and demands a baseline level of maintenance — but active pricing management of those SKUs produces minimal return.
SKU rationalization using price-performance data isn't about cutting products from your catalog. It's about identifying which SKUs deserve active, continuous pricing management and which ones can be put on auto-pilot with a floor-and-ceiling rule. The goal is concentrating analyst time and system resources on the decisions that actually move the margin needle.
What Price-Performance Data Tells You
Price-performance at the SKU level means answering two questions simultaneously: how does this SKU perform commercially (revenue, margin, sell-through velocity), and how does that performance correlate with its competitive pricing position? Both dimensions matter independently. But the intersection of the two is where the most useful signals live.
A high-revenue SKU that performs the same regardless of whether your price is 5% above or below competitors is price-inelastic in your catalog — you can probably set it at a premium and leave it there. A moderate-revenue SKU that sees meaningful volume swings when you're within 3% of competitor pricing is highly elastic and deserves active monitoring. A low-revenue SKU that no competitor actively monitors is probably fine on a cost-plus floor with annual review.
The analysis requires four data points per SKU over a trailing 90-180 day window: revenue contribution, gross margin percentage, sell-through rate, and a price elasticity estimate derived from comparing volume performance during periods at different price positions relative to your competitive set. That last one is the hardest to generate without tooling, but it's the most operationally useful.
Building the SKU Tier Framework
Once you have price-performance data, the natural output is a tiered framework that drives different management cadences and pricing rules per tier. Here's the three-tier structure that works well for catalogs in the 1,000-5,000 SKU range.
Tier 1: Active management (top 15-20% of SKUs)
These are your high-revenue, high-margin SKUs where competitive sensitivity is material. They warrant weekly or more frequent competitive monitoring, margin-delta ranking in your recommendation queue, and proactive review against demand velocity signals. A typical 2,000 SKU catalog will have 300-400 Tier 1 SKUs. These are the ones that justify investment in competitive intelligence tooling. Getting pricing right on this tier drives the overwhelming majority of margin outcomes.
Tier 2: Rule-based management (middle 40-50%)
Moderate contributors with some competitive sensitivity but not enough to justify daily analyst attention. These SKUs can be managed effectively with a simple rule set: price within a defined band relative to your top three competitors in the category, with floor and ceiling constraints. The rule runs automatically, and a human reviews the SKU only if it triggers an exception (price fell below floor, competitor exit, or demand velocity spike). Alert-driven rather than calendar-driven review.
Tier 3: Static with periodic review (bottom 35-40%)
Long-tail SKUs with minimal margin contribution and low competitive sensitivity. These are often highly specific variants (unusual size ranges, edge-case color options), niche accessories, or product categories where you have no meaningful competitive set. Set them at cost-plus floor, review quarterly or when a major catalog audit occurs, and don't allocate active analyst time. If a Tier 3 SKU starts moving, your velocity monitoring will surface it for reclassification.
The Analytics Pipeline for Tier Classification
Building this classification requires a few months of data and a deliberate analytical process. The most common failure mode is doing the classification once and treating it as permanent. SKU tier membership should be reviewed quarterly, because commercial performance and competitive dynamics both shift. A Tier 3 SKU that gets picked up by a major competitor for their own catalog can quickly become Tier 1 territory.
The classification inputs we recommend tracking per SKU over a 90-day trailing window: gross revenue, gross margin percentage, average daily unit velocity, standard deviation of unit velocity (a proxy for demand consistency), number of active competitors in your defined competitive set, average price gap versus competitive median, and price elasticity coefficient if you have A/B price test data or natural price variation to analyze.
For catalogs without historical price elasticity data, a reasonable approximation is competitive sensitivity score: how often has this SKU been within 5% of a competitor's price, and how did velocity correlate with those periods? High co-movement suggests elasticity. Low co-movement suggests inelasticity and probably a Tier 3 candidate regardless of revenue contribution.
A Concrete Illustration
A DTC outdoor gear brand carries 1,800 active SKUs across camping, hiking, and climbing categories. Their pricing analyst team of two was spending roughly equal time on each category. After a price-performance analysis, the picture looked like this: 280 SKUs (15.5% of catalog) in their technical climbing hardware and premium backpacking categories accounted for 83% of gross margin contribution. These SKUs had active competitive sets of 4-8 players and meaningful price elasticity. The remaining 1,520 SKUs — lower-priced accessories, apparel basics, climbing chalk, replacement parts — contributed 17% of margin with minimal competitive sensitivity.
After tiering, the two analysts focused 80% of their active management time on the 280 Tier 1 SKUs. The other 1,520 SKUs moved to rule-based or static management. Within two quarters, their recommendation throughput on the high-value tier increased substantially — not because they worked more hours, but because the noise from the long tail stopped competing for their attention. Margin improvement came not from adding resources but from concentrating existing ones.
We're not saying the long tail doesn't matter at all. Long-tail SKUs contribute to catalog completeness, average order value through complementary purchases, and customer retention in niche segments. The point isn't to cut these products — it's to stop treating them as if they require the same pricing attention as your top-margin drivers.
When to Reclassify a SKU
Reclassification triggers fall into three categories. An upward trigger moves a SKU from a lower tier to a higher one: meaningful velocity increase (more than 40% above trailing average for two consecutive weeks), new competitor entry into the category, or a gross margin spike that suddenly makes the SKU materially relevant. A downward trigger moves a SKU to a lower tier: sustained velocity decline, competitor exit leaving minimal competitive set, or a price inelasticity pattern becoming clear over a 90-day window.
A third trigger type is catalog structural change: new product introductions that cannibalize an existing SKU, seasonal product transitions, or bundle pricing changes that alter the economics of individual components. These require a more deliberate review rather than an algorithmic reclassification, since the context is qualitative as well as quantitative.
Operationalizing Tier Rules in Your Pricing Stack
Once you have tier classifications, the rules need to live somewhere actionable. For Tier 1, the rule is active monitoring plus analyst review of every significant recommendation. For Tier 2, it's a bounded automated rule with exception alerts routed to the analyst queue when bounds are breached. For Tier 3, it's a static price with an annual review flag.
In Orbivex, Tier 1 SKUs surface in the daily recommendation digest with full competitive context and margin delta. Tier 2 SKUs operate under a rule set you define (e.g., stay within 4% of category median, hard floor at cost-plus 55%) and only appear in the analyst queue when they breach their bounds or trigger a demand velocity alert. Tier 3 SKUs aren't actively monitored but are included in the periodic full-catalog review. The analyst workflow changes from "review 1,800 items" to "review 280 items actively, triage 400 exception alerts, schedule quarterly pass on the rest." That's a fundamentally different capacity equation.