AI Sales History helps AGR build a stronger demand history for items that have no sales history or limited sales history.

Using AI, AGR identifies comparable products in your assortment and uses their sales patterns to create an AI-generated sales history. This gives the forecasting engine more information to work with and can help reveal demand patterns that would otherwise be difficult to identify.

There are two main use cases:

And you don't just get an AI-generated result. AGR shows you which comparable items were used and the weight given to each one, so you can understand what the generated history is based on.

The goal is to give AGR a stronger demand signal for better forecasting, purchasing, and replenishment decisions.

The two concepts work together but serve different purposes.

AI Sales History creates the underlying sales history using comparable products. It can be used for both new items and items that already have some sales history.

New Item Forecast builds on that history to create a forecast for a new item with no history of its own. Instead of leaving planners with no forecast or requiring them to manually create a Sales Plan, AGR can use the item's attributes, price point, and comparable products to establish an initial demand expectation.

In both cases, the aim is not to present an unexplained AI-generated number. AGR gives you visibility into the information behind the result so you can review and understand it.

Forecasting relies on historical demand. When there is little or no history available, AGR has less information from which to identify the item's expected demand level and patterns.

For a new item, there is no historical sales pattern at all. This makes it difficult to determine how much stock to purchase or replenish at launch.

For an item with limited history, there may be enough data to understand its recent sales level but not enough to recognize longer-term patterns. For example, an item with only 9 months of sales has not yet completed a full annual cycle, making seasonality difficult to identify.

AI Sales History helps fill this gap by asking:

"Which products in my existing assortment are most comparable to this item, and what can their sales patterns tell us about its likely demand?"

AGR uses those comparable items to create additional sales history for the forecasting engine.

You can activate the AI sales history for individual items or use Bulk Update for multiple items.

Open an item and go to:

Item Info → Forecast Settings → AI Sales History

Enable the setting to generate the AI Sales History.

Here you can also review the comparable items AGR selected and the weight assigned to each one. Information buttons provide additional details, allowing you to understand how the generated history was created.

A useful workflow is:

You can disable AI Sales History again at any time.

For a new item with no sales history, AI Sales History creates a demand pattern that AGR can use as a starting point for forecasting and replenishment.

For an item with limited history, it extends the available demand picture further into the past, giving the forecasting engine more information to identify patterns such as seasonality.

Add the AI Sales History column or filter to your item list and filter for Yes.

This gives you an overview of all items currently using AI-generated sales

AGR uses an AI model built with Azure OpenAI to identify comparable products already in your assortment.

The comparison considers product information such as:

AGR searches for up to 10 comparable items within the same location that have usable sales history.

To keep the comparison as relevant as possible, AGR first searches within Item Group 2. If there are not enough suitable items available, the search is broadened:

Item Group 2 → Item Group 1 → All items within the location

Items that would not provide a useful demand signal, including special-order items and items without sales history, are excluded.

Each comparable item receives a similarity weight. Items that are more closely related have a greater influence on the generated history.

This means AGR does not simply copy the sales history of one product. It combines demand patterns from multiple comparable products according to how closely they match the item being analyzed.

Importantly, this process is visible to the planner. you can see which items were selected and their weighing, and use the available information buttons to understand more about how the result was created.

From these comparable items, AGR can generate up to 25 months of AI sales history reflecting both:

As real sales become available, AGR uses them to calibrate the AI-generated history.

After the item has completed at least one full forecasting period—one full month for monthly forecasting or one full week for weekly forecasting—AGR compares actual demand with the AI-generated baseline.

For example, if the comparable products suggest demand of approximately 50 units per week, but the new item sells 100 units during its first complete week, the generated history can be scaled to approximately 2× the original expected level.

This allows AGR to retain useful demand patterns from comparable products while adjusting the demand level to reflect how the item is actually performing.

As more real sales accumulate, the item's own demand history increasingly informs the forecast, allowing the forecast to evolve with actual performance.