Inventory and Demand Forecasting

The Inventory Tightrope

In the previous lesson, we turned customer reviews into business intelligence. Now let’s build on that foundation with the challenge that keeps every e-commerce seller awake at night: inventory management.

Too much inventory ties up cash and risks storage fees or dead stock. Too little means stockouts that cost you sales, search ranking, and customers.

The sweet spot is narrow. AI helps you find it.

The True Cost of Inventory Mistakes

Problem	Direct Cost	Hidden Cost
Stockout	Lost sales	Ranking drop, customer loss to competitor
Overstock	Storage fees	Cash tied up, potential write-offs
Wrong mix	Some items overstocked, others understocked	Opportunity cost of wrong inventory

Basic Demand Forecasting

Start with your historical sales data:

AI: Here are my monthly sales for the past 12 months for [product]:

[Month: Units sold]

Analyze this data and:
1. What's the average monthly demand?
2. Is there a growth or decline trend?
3. Are there seasonal patterns?
4. What would you forecast for the next 3 months?
5. What's the range of uncertainty (best case, worst case, likely case)?

The Moving Average Method

For products with relatively stable demand:

3-Month Moving Average = (Month 1 + Month 2 + Month 3) / 3

This smooths out short-term fluctuations and gives a reasonable baseline forecast.

Seasonal Adjustment

If your product has seasonal patterns (swimwear, holiday gifts, school supplies):

AI: Here are my weekly sales for the past 2 years:

[Paste data]

Identify:
1. Seasonal peaks and valleys
2. What percentage above/below average each month typically is
3. Week-over-week trends leading into peak seasons
4. A forecast for the next quarter that accounts for seasonality

The Reorder Point Formula

When should you place your next order?

Reorder Point = (Daily demand × Lead time) + Safety stock

Daily demand: Average units sold per day Lead time: Days from order to receiving inventory Safety stock: Buffer for unexpected demand spikes or supply delays

AI: Help me calculate reorder points for my product:

- Average daily sales: [X] units
- Supplier lead time: [X] days
- Variability in daily sales: [range]
- Variability in lead time: [range]
- Acceptable risk of stockout: [low/medium/high]

Calculate:
1. Minimum reorder point
2. Recommended safety stock
3. When I should place my next order
4. How many units to order

Quick Check

Your product sells an average of 10 units per day. Your supplier takes 14 days to deliver. You want 7 days of safety stock. What’s your reorder point?

See answer

Reorder Point = (10 units/day x 14 days lead time) + (10 units/day x 7 days safety stock) = 140 + 70 = 210 units. When your inventory hits 210 units, place your next order. The 140 covers the lead time; the 70 provides a buffer if demand spikes or delivery is delayed.

Advanced Forecasting Inputs

Beyond sales history, factor in these demand drivers:

External Factors:

Holidays and events: Black Friday, back-to-school, Valentine’s Day
Competitor launches: New competing products can shift demand
Market trends: Category growth or decline
Economic conditions: Consumer spending patterns

Internal Factors:

Promotional plans: Planned sales or marketing campaigns
Price changes: Price drops increase demand, raises decrease it
New product launches: May cannibalize existing products
Listing changes: Improved listings can boost conversion rates

AI: I'm planning inventory for Q4 (October-December) for my [product].

Historical Q4 sales: [data]
Planned promotions: [dates and details]
Known events: Black Friday, Cyber Monday, Christmas
Competitor changes: [any known]

Create a weekly demand forecast for Q4 that accounts for:
1. Historical seasonal patterns
2. Impact of planned promotions
3. Holiday buying spikes
4. A safety margin for uncertainty

Inventory Health Metrics

Track these numbers to keep inventory healthy:

Metric	Formula	Target
Inventory Turnover	Cost of Goods Sold / Average Inventory	Higher is better (varies by category)
Days of Supply	Current Inventory / Daily Sales	30-60 days for most products
Sell-Through Rate	Units Sold / Units Available	Track weekly during promotions
Dead Stock %	Unsold items > 6 months / Total Inventory	Under 5%

Managing Multiple Products

When you sell multiple items, prioritize inventory attention:

ABC Analysis:

A items (top 20%): Generate 80% of revenue. Monitor daily.
B items (next 30%): Generate 15% of revenue. Monitor weekly.
C items (bottom 50%): Generate 5% of revenue. Monitor monthly.

AI: Here are my product sales for the past quarter:

[Product: Revenue]

Classify these into A, B, and C categories and recommend:
1. Monitoring frequency for each category
2. Safety stock levels for each
3. Which products I might consider discontinuing
4. Where I should invest more inventory

Exercise: Build Your Inventory Plan

For your top product:

Gather 6-12 months of sales data
Forecast demand for the next 3 months with AI
Calculate your reorder point and safety stock
Identify seasonal or external factors to adjust for
Set up a monitoring schedule based on the ABC method

Key Takeaways

Stockouts cost more than the missed sale—they damage ranking, lose customers, and reduce future conversion
Demand forecasting combines historical data with seasonal trends and external factors
The reorder point formula (daily demand x lead time + safety stock) prevents both stockouts and overstock
ABC analysis helps you prioritize attention: top 20% of products get daily monitoring
AI turns complex forecasting math into accessible, actionable inventory plans
Track inventory health metrics monthly to catch problems before they become costly

Up next: In the next lesson, we’ll dive into Customer Personas and Targeting.