Why Your Forecasting Model Is Creating Overstock (And What To Do About It)

Every brand has a story about the season that went wrong. The buy was too deep, the trend shifted, the wholesale account cancelled, the weather didn't cooperate. These stories feel like bad luck — but the data tells a different story. Systematic overstock is almost always the result of structural problems in how brands plan and buy, not random misfortune.

The Scale of the Overstock Problem

Industry data consistently shows that consumer brands carry 15–25% more inventory than they can sell through primary channels at full price. For an apparel brand doing $100M in revenue, that's $15–25M of inventory that was bought and will never earn its full intended margin.

The downstream costs compound: carrying costs run 20–30% of inventory value annually (storage, insurance, cost of capital, obsolescence). A $20M overstock position costs $4–6M per year just to hold, before accounting for the discount or loss on eventual disposition.

Root Cause 1: Demand Forecast Bias

Most demand forecasting models are systematically optimistic. There are structural reasons for this:

• Anchoring on last year's numbers — Forecasters anchor heavily on prior-year performance and apply growth assumptions without adequately accounting for market shifts, competitive dynamics, or trend changes.
• Upside bias from sales teams — Sales and merchant teams, whose compensation is often tied to topline growth, tend to provide optimistic demand signals that inflate forecasts.
• Fear of stockouts — The pain of a stockout (lost sale, unhappy customer, wholesale chargeback) is immediate and visible. The pain of overstock (slow-moving inventory, eventual markdown) is diffuse and delayed. This asymmetry pushes forecasters to buy more rather than less.
• Failure to account for cannibalization — New style introductions often cannibalize existing styles more than planned. Total category demand stays flat while unit count increases.

Root Cause 2: Incentive Misalignment in Buying

Buyers are typically measured on in-season sell-through and gross margin, not on end-of-season inventory position. This creates a predictable dynamic:

• Buyers optimize for in-season success — they buy deep on styles they believe in
• They are not penalized for overstock that shows up in the secondary channel months later
• The secondary team inherits the problem but had no input into the buy

Brands that have successfully reduced structural overstock typically do one thing: they include a "secondary recovery cost" line in buyer performance measurement. When buyers see that their overstock costs them 15 cents on the dollar in secondary disposition costs, they start buying more conservatively.

Root Cause 3: Safety Stock Miscalculation

Safety stock — the buffer inventory held to cover demand variability and supply uncertainty — is frequently calculated too high. Common mistakes:

• Using maximum lead time rather than average lead time in safety stock calculations
• Applying safety stock uniformly across all SKUs regardless of velocity
• Not adjusting safety stock as supply chains become more reliable and lead times shorten
• Counting safety stock at the SKU level rather than at the size/color level, masking size distortions

A brand that reduces its safety stock calculation from 8 weeks to 5 weeks — a defensible change given modern supply chain visibility — can reduce inventory investment by 10–15% without meaningful increase in stockout risk.

Root Cause 4: The Open-to-Buy Ratchet

Open-to-buy (OTB) is the budget available for new inventory purchases in a given period. In many brands, OTB is calculated without adequately crediting the inventory already on hand or on order. The result: each planning cycle starts with an inflated budget that buys more than needed.

The fix is disciplined OTB management that fully accounts for current inventory position, projected sell-through, and committed inbound. This sounds obvious — but it requires clean inventory data that many brands don't have.

Making Secondary a Planned Channel, Not an Emergency Valve

The most sophisticated brands treat secondary inventory not as a symptom of planning failure, but as a planned part of their inventory strategy. This means:

1. Build secondary budgets into the buy plan. If you expect 15% of your inventory to flow to secondary, plan for it. Build a recovery assumption into your margin model.
2. Plan secondary channel capacity in advance. If you know you'll have a sample sale in November, start planning in September. Don't scramble to organize an event with three weeks' notice.
3. Track secondary recovery as a KPI alongside primary sell-through. The best brands review both metrics together in monthly business reviews.
4. Feed secondary data back into the buy. Which styles consistently end up in secondary? Which categories carry the highest overstock rates? This data should directly inform next season's buy plan.

AI-Driven Forecasting: Promise and Limitations

AI-driven demand forecasting tools have improved significantly and can meaningfully reduce forecast error — particularly for brands with rich historical data and stable product lines. What they can't do:

• Predict the impact of new trend shifts not present in historical data
• Override the human incentive structures that drive overbuy behavior
• Compensate for bad input data (garbage in, garbage out)

AI forecasting is a valuable tool, but it works best when it's one input into a disciplined planning process — not a replacement for it.