OrderIQ is installed and configured for the customer account
​
​

Customer has at least 2 years of purchase history for best results (minimum required)
​
​

Products must be active and have SalesReportDefault = 1 on the Product Type
​
​

Customer must have an active route with scheduled stops

Previous 8 weeks of individual purchase quantities
​
​

Rolling averages over 4, 9, 13, and 26-week periods
​
​

Purchase frequency and weeks between purchases
​
​

Seasonal patterns (year, quarter, month, week)
​
​

Weather conditions (temperature and precipitation)
​
​

Pricing changes and discounts
​
​

Special events and holidays (when configured)
​
​

POS data if enabled (retail inventory and sales velocity)

Apply: Must be enabled for the customer to receive OrderIQ suggestions
​
​

AISuggested Order: Must be enabled to use the AI-powered prediction model
​
​

Order Size: Defines how to split the weekly forecast across multiple delivery stops (e.g., "0.5;0.5" splits evenly across two stops)
​
​

Round Threshold: Decimal threshold for rounding up (e.g., 0.5 means quantities ≥ 0.5 round up, < 0.5 round down)
​
​

Force To Case: When enabled, ensures suggestions are always in full case quantities
​
​

Enable Layer Round: Rounds quantities to match pallet layer configurations
​
​

Enable POS Data: When enabled, incorporates point-of-sale data for more accurate predictions

Allow at least 4-8 weeks of data collection before expecting accurate predictions for new customer/product combinations
​
​

Products with sporadic purchase history (long gaps between purchases) may have less accurate predictions initially
​
​

Review and adjust suggestions for the first few weeks to help the system learn actual ordering patterns
​
​

Configure Special Events for holidays or promotions that significantly impact sales
​
​

If a product consistently over- or under-predicts, check if the product's purchase pattern is unusual (seasonal, event-driven, etc.)