Inventory planning with 60-day ocean lead times & 100+ SKUs – how do you control forecasting?

We tested this: a one‑week stockout on a stable top‑selling SKU cost us 5x more in lost profit than holding an extra month of inventory. So we now bias toward a little too much rather than too little for our core ASINs. For non‑core items, we bias toward lean.
Also, never let slow SKUs linger. If something hasn't sold well for 3‑6 months, liquidate or dispose. The cost of holding dead stock is higher than most people realize.

on manager approval and deviation analysis**,Operators submit forecasts, but a manager must approve any forecast that deviates more than 20% from the weighted average (unless there's a clear reason, like a planned promotion). The manager's job is to question the outliers.
After the fact, run a forecast accuracy report. Compare actual sales to the forecast from 2‑3 months ago. Hold a monthly "bias review" meeting. Don't punish errors – punish not learning from them. If someone is consistently off by 30% without justification, that's a coaching opportunity.

We require operators to update daily sales, remaining stock, and expected sell‑out dates for each SKU. They also track in‑transit inventory. This data is reviewed by management daily – not to micromanage, but to spot risks early.
Once a week, we hold a 30‑minute meeting to review:

• SKUs at risk of stockout (reorder urgently)
• SKUs with too much inventory (adjust pricing, slow down orders)
• Any changes in the market (competitor stockouts, price drops, new entrants)

This turns forecasting from a monthly gamble into a weekly adjustment process.

Categorize your 100+ SKUs:

• Stable best‑sellers – protect these at all costs. Order confidently, keep safety stock.
• Profitable but lower volume – order reasonably, but don't overcommit.
• New/potential products – order conservatively. If they show traction, you can ask suppliers to hold raw materials (not finished goods). Raw materials cost less to carry and can be turned into finished goods faster.
• Slow or unstable – cut order quantities or stop. Don't throw good money after bad.

Also, work with suppliers to shorten lead times. 30 days production + 60 days shipping is too long. Aim for 15 days production. That's a strategic shift, but it's the only way to reduce risk.

on dynamic monitoring and lifecycle management,Forecasts will never be perfect, but you can catch problems early.

• Track weekly sales vs. the previous 2‑3 weeks. If sales drop 20% or more, pause or reduce next orders.
• If sales jump 20%+, expedite production or switch to faster shipping.
• Use product lifecycle stages: new products, growth, mature, decline. Adjust ordering accordingly. Most over‑ordering happens with new products and slow‑moving tails – cut those aggressively.

Also, if you have historical data, adjust for anomalies. For example, last year's Q2 might have been affected by a market event (e.g., war, supply chain crisis). Don't blindly copy old numbers.