Skip to main content
All CollectionsInventory Optimization
Inventory: Demand Forecast
Inventory: Demand Forecast
Updated over 4 months ago

Planning for the future sales forecast can be tricky and time consuming. Flywheel’s Demand Forecast model is always running and constantly adjusting to help save you time and energy in monitoring your product’s inventory availability and projected remaining days of inventory.

Demand Forecast - Watch Video

Flywheel is connected to your merchant accounts and receives updates throughout the day of your current FBA / WFS inventory levels - and merchants generally recommend at least 30 days of active available inventory on hand. Recommended restock actions from the merchant calculations also tend to be delayed which may be misleading.

The Demand Forecast page breaks down for all active products, how many days of inventory are remaining based on Flywheel’s AI calculated estimated stockout date using not only the calculated current Rate of Sale from the past 30 days; but incorporates your past sales data for 12+ months, identified anticipated seasonal demand dips or spikes based both your historical seasonal trends as well as indexed values from our benchmarked data to predict the future demand at the SKU level.

Data and prediction models will change estimations weekly as realized sales and actual inventory fluctuate impacting predicted demand up to 365 days reflected in the Units Projected to Sell. The model anticipates future sales with an impact of seasonal trends, so the Days of Inventory and Projected Units may fluctuate.

Your displayed Rate of Sale indicates how many units on average have been sold in the past 30 days; but the reflected remaining Days of Inventory now calculates a unique value that may not be a strict division of inventory available and rate of sale.

For example:

  • It is October and you have 100 Santa hats with a Rate of Sale averaging 1 unit / day the past 30 days

  • In this scenario, the Days of Inventory may reflect 25 days because our models show that historically, sales spike in November ahead of the Christmas seasonal demand; and we expect the 100 units to sell out in 25 days because expected ROS will increase to 4/day

    • The displayed value doesn't strictly look at 100 hats / 1 unit lasting 100 days based on your past 30 days of sales; but in combination with the seasonal peaks expected in November

  • If it was January, that same calculation may instead reflect 160 days of inventory remaining because the rate of sale in a non-seasonal demand period to drop to <1 unit/day

Review the projected inventory levels and think about items requiring restock, advertising adjustments or potentially liquidation opportunities to help adjust inventory to support predicted demand.

Demand Forecast Applications:

  • High Demand Forecast:

    • Increase Ad Spend: If the forecast predicts high demand, consider increasing ad spend to capture the anticipated market interest.

    • Promotional Campaigns: Launch targeted promotional campaigns to generate buzz and drive sales.

    • Focus on High-Converting Channels: Allocate more budget to channels that historically have high conversion rates.

  • Low Demand Forecast:

    • Cost Efficiency: Optimize ad spend by focusing on cost-effective channels and ad formats.

    • Retention Campaigns: Run retention campaigns to keep existing customers engaged and loyal.

    • Target Niche Audiences: Identify and target niche audiences that may have a higher interest in the product despite overall lower demand.

Download capability is available for CSV reports for further offline analysis, or if you would like to create a custom notification for an email alert when items meet a specified Days of Inventory remaining to help prevent stockouts or a reminder at a specific reorder level, or dive into specific Rate of Sale changes

Did this answer your question?