The Scale of the Electrical SKU Problem
Electrical distribution is among the most catalog-intensive segments in all of distribution. A mid-market electrical distributor serving commercial, industrial, and residential contractors might carry 30,000-80,000 active SKUs across wire and cable, conduit and fittings, wiring devices, panelboards, transformers, lighting, and automation products. Each of those product families has its own specification matrix - gauge, ampacity, voltage rating, conduit type, conduit fill, code compliance certifications - that creates branching complexity.
Industry data confirms the scope: 70% of electrical distributors manage 5,000+ active SKUs, and a meaningful segment is operating at 50,000+. The problem is not just the number of SKUs - it is that demand for specific SKUs is highly volatile and project-driven. A medium-sized commercial electrical contractor winning a large hospital project can immediately spike demand for specific wire configurations that represent negligible volume in normal periods.
The result is a persistent inventory paradox: stockouts on the exact SKUs that are suddenly hot, while deadstock accumulates on SKUs that looked reasonable to stock when the forecast was built. Industry surveys put deadstock at 6-10% of inventory value for 26% of electrical distributors. At a $20M inventory value, that is $1.2-$2.0 million in cash tied up in product that is not moving.
Why Min-Max Systems Fail at This Scale
The ERP-based min-max inventory systems most electrical distributors rely on were designed for a world where demand variation was modest and predictable. Set a minimum (reorder point) and a maximum (order-up-to level) for each SKU based on historical velocity and lead time, and the system generates purchase orders automatically when stock hits the minimum.
This works acceptably for high-velocity commodity SKUs with stable demand - standard wire gauges, common conduit fittings, frequently specified devices. It fails badly on three categories of SKUs that are increasingly common in electrical distribution:
- Project-spiked SKUs: Products that see near-zero demand for 6-8 months and then explosive demand when a large project hits your service area. The historical average suggests low-velocity; the actual demand pattern is feast-or-famine. Min-max systems stock for the average and miss the feast.
- Code-transition SKUs: Products affected by NEC code updates, energy efficiency mandates, or utility rebate programs can shift from slow-movers to fast-movers overnight when a code cycle takes effect. Min-max systems cannot anticipate regulatory change.
- End-of-life SKUs: Products being phased out by manufacturers accumulate as deadstock when distributors do not have visibility into manufacturer lifecycle announcements. Min-max keeps reordering product that is becoming harder to sell.
AI Demand Forecasting: What Changes
AI demand forecasting replaces the historical-average assumption with a forward-looking model that incorporates multiple signal types. For electrical distribution, the most predictive external signals are permit pull data (a leading indicator of project-driven demand 8-16 weeks out), project award databases (Dodge Data, Construct Connect), utility rebate program calendars (a known demand driver for lighting and efficiency products), and regional economic indicators (commercial construction starts, industrial capacity utilization).
Combined with internal signals - account-level purchase patterns, quote activity that has not yet converted to orders, seasonal adjustment factors - an AI demand model generates SKU-level forecasts that are materially more accurate than min-max for the problem SKU categories listed above.
The operational impact: instead of a static min-max parameter that does not change unless someone manually adjusts it, the system generates dynamic reorder parameters that reflect current demand signals. For a high-velocity commodity SKU, this might mean modest adjustments week to week. For a project-spiked SKU where permit data shows a major project coming, the system flags the increase 8-12 weeks before demand hits - giving procurement time to position inventory without emergency expediting.
SKU Rationalization: The Hardest Conversation
AI demand forecasting improves how you manage your existing SKU catalog. SKU rationalization addresses whether your catalog is the right catalog. The two work together: better demand data makes rationalization decisions easier to defend with data, and a smaller, cleaner catalog makes demand forecasting more accurate because the model is not processing noise from SKUs with near-zero movement.
The rationalization framework for electrical distributors:
- Velocity analysis: What percentage of your SKUs drive 80% of your sales? For most electrical distributors, 20-25% of active SKUs generate 80%+ of revenue. The long tail needs scrutiny.
- Carrying cost calculation: For each slow-moving SKU, calculate annual carrying cost (capital cost + storage cost + handling cost + obsolescence risk). Compare to annual gross profit contribution. SKUs where carrying cost exceeds gross profit contribution are candidates for elimination or special-order conversion.
- Account dependency mapping: Before eliminating any SKU, check whether it appears in the purchase history of high-value accounts. A SKU with low aggregate velocity that is disproportionately important to a $200,000/year account has very different economics than a low-velocity SKU with no strong account attachment.
- Substitutability assessment: For each rationalization candidate, identify the closest stocking SKU that could serve the same application. If a stocking substitute exists with acceptable spec overlap, elimination is lower risk.
The Warehouse Automation Connection
The 45% of electrical distributors embedding automated warehouse solutions are building something beyond operational efficiency - they are building a data asset. Automated systems capture granular movement data (every pick, every replenishment, every no-movement period) at a level of precision that manual warehouses cannot match.
This movement data feeds directly into SKU performance models. A carousel that records the last pick date for every location gives you a real-time deadstock map. An AS/RS system that logs slotting history reveals which SKUs have been relocated for slow-movement - a signal that triggers rationalization review.
Distributors who have integrated their automated warehouse data with AI demand forecasting report the compound effect: better forecasts reduce the inventory that needs to be managed in the first place, while warehouse automation makes managing the remaining inventory more efficient. The combination is what moves the needle on the deadstock percentage from 6-10% down to 2-4% without sacrificing fill rates.
All Tools