Key Concepts and Terminologies

Regular Price

The base, non-promotional selling price of an item at a location or price zone. Regular Pricing Optimization in LPO recommends regular prices that align with configured business rules, pricing objectives, and market conditions.

Price Zone

A price zone is a logical grouping of stores or online locations that share the same pricing. LPO can generate regular price recommendations at the price zone level, applied uniformly across all stores in the zone or at individual location nodes in the location hierarchy, such as region or store group. A single LPO instance cannot generate recommendations simultaneously at both price zone and location node levels.

Merchandise Hierarchy

Merchandising Hierarchy is the product classification structure used in LPO to scope pricing rules and runs. Rules can be defined at any level of the hierarchy for example, Department, Class, Subclass, Style, or Style/Color, and LPO applies them to all items within that node. Runs are typically configured at a higher hierarchy level (for example, Department), while recommendations are generated at a more granular level (for example, Style/Color).

Non-Forecast (N/F) vs Forecast (F) Rules

In LPO, every Regular Pricing rule is classified as either Forecast-based (F) or Non-Forecast-based (N/F).

N/F rules operate on current data such as current prices, cost, competitor prices, and so on. They do not require demand forecasting. All rules used in Rules Based Regular Pricing are N/F rules.
F rules require a demand forecast to calculate expected revenue, volume, or margin impact. These rules are not available in Rules Based Regular Pricing. If a strategy contains F rules, they are ignored during a Rules Based run.

Business Rules and Strategy

A strategy determines which Non-Forecast (N/F) rules are evaluated during a Rules Based Regular Pricing run. Strategies are discussed in detail in Chapter 5 - Rules and strategies .

Strategies can be scoped to specific merchandise levels and locations. Every LPO instance has a system-provided default strategy (DEFAULT_SET) used for scheduled batch runs. You can also create What-If strategies to test alternate rule combinations to address specific business needs, such as high-priority products, promotional events, or competitor activity without affecting live recommendations.

For additional details on business rules and strategy, see the Control and Tactical Center chapter in the Oracle Retail AI Foundation Cloud Service User Guide.

Batch Run Vs What-If Run

Every regular price recommendation generated by Rules Based Regular Pricing originates from a run. A run defines the business problem that LPO is trying to solve by combining a specific scope and strategy. Depending on the business need, runs can be executed as scheduled batch runs or user-created what-if scenarios.

Batch Run

A batch run is a scheduled, automated execution of the optimization engine using the default strategy. Results from the LPO batch runs are automatically moved to the Manage Recommendations workspace for analyst review and approval.

For Regular Pricing, recommendations from a successful batch run are automatically approved and exported to downstream systems if they meet auto-approval criteria configured in Manage System Configurations. These criteria evaluate whether:

The new recommended price falls below a specified threshold.
The percentage change in price is within a defined limit.
The absolute difference between the old and new price is within a predefined amount.

Regular price recommendations are also subject to export frequency rules, which control how often an item can receive a new recommendation. These rules ensure pricing changes happen at appropriate intervals and include conditions such as:

A minimum number of days must have passed since the item's last regular price change.
Recommendations are only made on specified days of the week, for example, weekdays with lower shopping frequency such as Tuesday.
A minimum number of days must have elapsed since the end of the item's most recent promotional price period.
A minimum lead time is required between regular price changes before a new recommendation can be executed.

What-If Run

A What-If run is a user-initiated, adhoc execution that uses a custom pricing strategy. It enables pricing analysts and managers to evaluate the impact of alternative rule configurations before applying them to production.

Unlike scheduled batch runs, a What-If run does not replace the active recommendation set unless it is explicitly finalized.

You can use a What-If run to evaluate scenarios such as:

Testing a new competitor pricing rule that narrows the allowed price gap with a key competitor.
Increasing the minimum list margin for a product category to assess its impact on recommended prices.
Comparing the recommendations generated by two different pricing strategies.
Evaluating changes to price ladder or inter-item pricing rules before updating the default strategy.

Because a What-If run is isolated from production, you can compare its recommendations with those of the current batch run or with other What-If runs before deciding whether to promote the strategy for operational use.

Pricing Recommendations

Rules Based Regular Pricing generates regular price recommendations through scheduled Batch runs and user-initiated What-If runs. At a high level, this process consists of two major stages:

Run Overview, where runs are created, managed, and monitored.
Run Optimization, where LPO evaluates pricing opportunities and generates recommendations.

Run Overview

Run Overview is the central workspace for managing optimization runs in LPO.

Users typically access Run Overview to:

Create new optimization runs.
Monitor optimization progress.
Review completed runs.

For more details, see Run Overview and Management.

Run Optimization

Run Optimization is the process through which LPO evaluates pricing opportunities and generates recommendations. During optimization, the application evaluates the candidate regular prices against the applicable Non-Forecast pricing rules, resolves conflicts based on rule type and priority, and determines the Best Feasible Solution for each item in scope.

Although optimization is typically configured at higher merchandise and location levels, such as Department, Region, or Price Zone, recommendations are often generated at a more granular level, such as Style/Color, Item, Location, or Customer Segment.

For example, a retailer may create a run for the Women’s Activewear department within a price zone. During optimization, the application evaluates current prices, costs, competitor prices, and applicable Non-Forecast pricing rules before generating item-level regular price recommendations.

The optimization engine evaluates candidate regular prices against the applicable Non-Forecast pricing rules and constraints. When all rules cannot be satisfied simultaneously, it resolves conflicts using rule type and priority and returns the Best Feasible Solution as the recommended regular price.

Manage Recommendations

Manage Recommendations is the workspace where users review, approve, reject, and override price recommendations after a run completes.

For details, see Chapter 8 Manage Recommendations.

2 Key Concepts and Terminologies