Forecasts learned from real demand, not from empty shelves
Veritico STOCK forecasts demand for every item and store or warehouse, day by day. Before it calculates, it cleans the history of stockouts, outliers and promotion effects, then picks the method with the lowest expected error from around 80 statistical models. Planners adjust the forecast where they know more and confirm it.
Why forecasts miss
Learning from lost sales
When an item was out of stock, the history shows no sales. A forecast built on that history orders even less the next time.
One formula for every item
Fast movers, sporadic items and seasonal goods need different models. A single spreadsheet formula fits none of them well.
Late reaction
Weather, competitors or a holiday shift demand within days, while a monthly plan catches up only after the stock is gone or piled up.
How it works
How Veritico STOCK handles it
01
Clean history
Stockouts, outliers and promotion effects are removed, so the system learns what customers would have bought.
02
Choose the method
Around 80 statistical methods are tested per item, the one with the lowest expected error is used and its parameters are tuned per item.
03
Split into days
The monthly forecast is split into days per item and location, with seasonality and moving holidays.
04
Adjust and confirm
Planners adjust the forecast where they have extra information, confirm it and can freeze it for a period.
What you set in Veritico STOCK
Forecast level
Bottom‑up, top‑down or middle‑out. Middle‑out helps with new and strongly seasonal items.
Seasonality
Seasonal profiles including moving holidays. Extreme seasons are recognized and false seasonality is suppressed.
Demand sensing
The last few weeks are compared with long‑term history, and the forecast adjusts when they diverge.
New products
A new item takes over a set share of its predecessor’s history.
Promotion detection
Promotions are detected in the data even where nobody flagged them, so they do not distort the baseline.
Results with Veritico STOCK
Statistical forecasting methods, chosen per item
~0
Availability at Zeelandia
+0%
Veritico STOCK in practice
Inventory optimization for GGT
Automated forecasting and replenishment across 800+ branches delivered higher sales, leaner stock and faster ordering.
+5% revenue
–15% stock
–60% time to create orders
Seasonal inventory optimization for Zeelandia
Veritico STOCK helps Zeelandia manage strong seasonal demand, increasing product availability by 9% while cutting working capital in inventory by 18% and shortening turnover time by 36% to 70 days. Transparent, forecast‑driven planning now connects purchasing, production, warehouse transfers and distribution in one process.
+9% product availability
−18% working capital in inventory
−36% inventory turnover time, down to 70 days
Related solutions
The Veritico STOCK module: forecasting, stock targets and automated orders.
Promotion forecasts per store and day.
Forecasts for items without history and seasons that move.
Questions about demand forecasting
It estimates the sales lost on days when an item was unavailable. The forecast then reflects real demand, not what happened to be on the shelf.
Around 80 statistical methods combined with machine learning and neural networks. The system automatically picks the one with the lowest expected error for each item.
Yes. Planners can adjust the forecast, confirm it and freeze it for a chosen period.
Book a Veritico STOCK demo
Tell us how you order today, how many items and locations you manage and which KPI matters most. We’ll show you Veritico STOCK on a similar setup.