Frequently Asked Questions

This is a collection of frequently asked questions from our users.

How can I analyze the results of an optimization?

You can check the details of the optimization by clicking on the "eye" icon.

If this is one of the first optimizations you want to upload, it makes sense to check generally if the data seems valid to you. That can include checking the validity of current prices, product attributes, as well as red price discount and voucher levels.

If you are already familiar with optimizations in general and established data quality, it is often useful to analyze the most extreme price predictions by ordering the optimization results by price change, as well as checking the recommendations for high sellers, using the profit and revenue prediction filters. If you want to understand better what led to the price recommendation, you can check the diagram in the "Explain"-column, which will give some context for the recommendation. Lastly, it can be useful to check if there are products at negative margins, and if a price increase is recommended accordingly.

How long are optimization results available?

Old optimizations, comparison tables and uploads are available for two years in the front-end. Forecasts are only available for 3 month, as they take-up a lot of space. To increase data retention times, we can set-up a back-up flow into a cloud bucket in the upload section to save data for more than 2 years.

How do you create an optimization scenario with single or multiple targets?

Access the "Optimizations" page within your domain and click "Create." Provide a name, select the latest forecast, and define the scope for your target. For multiple targets, click "+" to add more targets and specify their respective scopes. You can include pre-defined rules for each target, such as "current price" or "interpolated price" rules. Finally, click "Submit" to create the optimization.

When is the new forecast going to be ready?

The new forecast is ready as soon as our machine learning pipeline to update the forecast ran through. Our pipeline is scheduled to start every day at the exact same time and includes the following steps:

1. We check the quality of your new data -> data quality issues will delay the pipeline
2. We train new models based on your new data and test them -> test errors will require a 7L Data Scientist to check
3. We create a new prediction with the new models and test it -> test errors will require a 7L Data Scientist to check

How can I compare several optimizations?

The compare dashboard enables side-by-side analysis of different scenarios or data sets. For this, simply click the "eye" on more than one optimization at a time. Apply filters based on specific criteria or variables to refine your comparison and focus on relevant information.

How can I schedule an optimization upload?

To schedule an upload, go to the "Upload" page and click the "Create" button. Set the desired upload time (e.g., after 2 minutes). After a short period, you'll receive an email notification requesting approval for the pending upload. Click the link in the email and confirm the upload. The status on the platform will update to reflect the successful upload, and the data will be available for use in optimizations.

Why are some optimized prices outside the requested price range?

The current price is always considered as a potential price point. For example, even if you requested to limit price points below the recommended retail price (RRP), the optimal price may still exceed the RRP if the current price is higher. In this scenario, to ensure that the optimal price remains strictly below the RRP, apply the following rule:

Optimal Price ≤ RRP × 100%

This rule guarantees that the optimal price will never exceed the RRP even if the current price is higher. You can find more information in section Rules.

How can I download optimization data?

To download optimization data, click the upload/download icon located to the right of the eye icon in the Optimization Overview. This will allow you to download a CSV file of the Comparison Table.

How can I get support when using the Pricing Tool?

On the right-hand side of the Pricing Tool, you will find a Help Desk option. From there, you can create a support ticket for our team.

When submitting a ticket, please choose the category that best fits your issue — Bug, Question, or Feature Request — and provide a clear and detailed description. You can also attach screenshots to help us understand the context.

Our team will respond as quickly as possible, and you will receive a notification once we have replied.

How and when can I re-run the pipeline?

You can re-run the pipeline yourself whenever the tool notifies you of a data issue that needs to be fixed. After resolving the issue, you do not need to wait for a data scientist to re-run it for you. Simply go to the Pipeline Status option on the right-hand side of the tool and trigger the re-run from there.

How is the optimal price chosen?

In the Optimization Comparison Overview, under Pricing+, there is a field called "Why this price?". This field explains how the tool arrived at the optimal price for each market–channel–product combination (or other configured levels of granularity).

When you click the icon in this field, you will see:

• The current price
• The list of allowed prices configured for that product
• The pricing rules applied to the product
• How each rule affected the allowed prices to determine the final optimal price

How pricing rules are applied:

• Rules are applied in priority order, from top to bottom
• Higher-priority rules are applied first
• Optimization starts with all allowed prices for the product
• Each rule filters the remaining allowed prices based on its conditions
• The output of one rule becomes the input for the next rule

What happens if a rule removes all prices:

• If a rule filters out all allowed prices, the rule is marked as failed
• Failed rules can be viewed in the Optimization Dashboard
• A fallback rule is then applied if configured

Final price selection: After all applicable rules are processed, the optimization logic evaluates the remaining prices using predicted profit and the configured steering target to select the optimal price.

Why are the rules not applied as I expected?

There can be several reasons why a rule is not applied:

1. Contradicting rules – Two or more rules conflict with each other. You can quickly identify such “failing rules” by filtering for failing rules in the filter menu.
2. Price option rejected by a rule for one product in the optimization group – In this case, that price option is immediately removed and cannot be used for any product in the group.
3. Allowed price range is not sufficient – The range is narrow to set the desired price or apply the rule.

How can I expand the list of allowed price points available for optimization?

There are two ways to introduce additional price points:

1. Update the allowed price points: if you need to expand the set of allowed price points, please create a support ticket via the Help Desk so our team can update them for you.

2. Use a price-generating rule: rules that use the “=” operator can generate new price points. For example, you can create a rule that sets the price to Purchase Price × X. This will produce a new price point, which will still follow any rounding rules defined for your allowed price points.

Can I use one rule across several different scopes without creating duplicates?

Yes! There are two ways you can achieve the desired flexibility:

1. Adjust the scope of a rule directly in the optimization definition.
   When creating an optimization, simply click on a rule and select the scope you want to use. This makes it easy to apply the same rule across different scopes.

2. Group rules together using the grouping option on the right-hand side of the Rule when creating an optimization.
   This allows you to combine multiple rules—such as the same rule template applied to different scopes—and use them together within an optimization.

How does the tool handle a product that belongs to multiple scopes?

If a product appears in more than one scope, the tool will display a “scope overlap detected” warning when you run an optimization. In this case, the tool optimizes the smaller (more specific) scope first, and the larger scope will not influence the result for that product.

What is the Optimization Group Identifier in the optimization comparison view?

An Optimization Group is a feature that treats a specific set of products as a single unit during price optimization. It can be a combination of market, channel, product group id, or product id It ensures that the prices of related products move together. The system determines the optimal relative price change (e.g., a 10% decrease) for the entire group, preserving the established price relationships (the price gap) between the items.

How does the pricing work inside an Optimization Group?

The optimizer decides on a single, uniform relative price change for the whole group.

The price difference (e.g., Product B is always twice the price of Product A) is maintained, but all prices shift up or down together based on the optimal decision for the group.

How do my pricing rules affect products within an Optimization Group?

Rules are tested individually for every product, but they impact the entire group's decision.

If a specific price option (e.g., increasing the price by +10%) is rejected by a rule for just one product in the group, that price option is immediately removed and cannot be used for any product in the group.

The final, optimal price change chosen must be applicable to all products in the group without violating any individual product's rules.

How can I include competitor price data in the pricing tool?

You can provide competitor price data as part of the daily import together with your other data tables. The required format and structure are described in Section 3.9, where you will find the full specification for the competitor price table.

Once the data is provided, you can use the existing rules to incorporate competitor prices into your optimization:

• Optimal Price ≥ Min Comp Price × value
• Optimal Price ≤ Min Comp Price × value
• Optimal Price = Min Comp Price × value

These rules allow you to anchor your optimized prices to competitor pricing as needed.

How does the system determine the initial price or discount for a new product with no sales history?

Initially, you can set a starting price and purchase price for the new product. Because the item lacks historical transaction data, the system determines its pricing behavior through a two-stage process:

• Cross-Learning: The system identifies "comparable" items already in the system by looking at attributes such as category, season, and brand. It initially assumes that costs and elasticity are similar to these products.

• Direct Learning: As the product generates incoming transactions, the model learns from it directly. It then adjusts the initial assumptions to reflect the product's actual performance.

The exact attributes influencing this cross-learning vary by model. If you have insights into specific factors that influence your products' elasticity, costs, or sales, please let us know so we can refine the model.

Can I prevent new products from being discounted for a specific period after launch?

Yes. You can define rules such as "Discount = 0%" for specific scopes (e.g., "New Arrivals") to ensure they remain at the Recommended Retail Price (RRP).

If a "new product" flag is not already available in your data, you can add a custom column to your Product Attributes table (e.g. FLAG_NEW). Using this attribute, you can create a specific scope in the app to maintain a price stability for these items by applying a fixed price or discount constraint.

If a specific rule you require for new products is not available in the tool, you can create a ticket requesting the feature and we will add the rule for you.