What is Selleasy AI and how it can be used?

Why choose Selleasy AI instead of manually created campaigns?

Save time and effort

Meaningful recommendations powered by ChatGPT

Better conversions with an AI upsell pitch

Auto-rotate non-performing campaigns

1. Setting Up a Selleasy AI Campaign

• Select the AI Option:
• Navigate to the Offers section in your Selleasy app.
• Choose the “ChatGPT powered Selleasy AI” option for your campaign.

• Set the Maximum Number of Products:
• In the campaign settings, you can specify the maximum number of products that can appear as recommendations.

2. Recommended Product Types

• Complementary Products:
• These are products that complement the trigger product or act as accessories.
• They are designed to enhance the customer’s purchase experience by pairing well with the main product.
• Example: If the trigger product is a blue shirt, Selleasy AI might recommend matching caps, shoes, or trousers.

• Similar Products:
• These are products that closely resemble the trigger product in style or type.
• They allow customers to explore alternatives while staying close to the original choice.
• Example: For a blue shirt, AI may suggest similar shirts in different colors or styles.

3. Providing Instructions to the AI (Optional)

Instruct AI on how to search for recommended products

• Example: "Only include products from Joggers collection"

Instruct AI on how to sort the search results

• Example: "Rank the lowest priced products at the top."

4. Filtering the Recommendations

• Enable the “Filter search results” option to start setting up filters.

• Recommendations can be filtered based on:
• Collection
• Tag
• Product Type
• Vendor

• Adding multiple values in a single condition: More than one collection, tag, product type, or vendor can be added within the same condition.
• Only for Collection and Tag, when multiple values are added, you can choose how they should be matched:
• “in any of the collections/tags” – products matching at least one of the added values will be shown.
• “in all collections/tags” – products matching all of the added values will be shown.
• Example: In the screenshot above, the condition is set to “in all collections” with Black and T-Shirts added. This will show only the recommended products that belong to both the Black and the T-Shirts collections.

• Adding multiple conditions: Click “Add condition” to create additional filter conditions.
• When more than one condition exists, you can decide how they combine:
• “All conditions must match” – a product must satisfy every condition to be shown.
• “Any condition must match” – a product satisfying at least one condition will be shown.

5. Deciding Recommendation Mode and Search Method

• Fast model (default) – Consumes 1 AI credit per recommendation. Fast model recommendations are generated quickly and are satisfactory for most use cases.

• Thinking model – Consumes 2 AI credits per recommendation. This model takes more time to generate recommendations, as it analyzes the products more thoroughly.

• GraphQL product search (default) – This is the default and most commonly used search method.

• Storefront MCP – If the GraphQL product search is not working for some reason, you can switch to the Storefront MCP option and try again.

6. Generating an Upsell Pitch

• The pitch content is automatically generated by AI which will be shown in the widget on the storefront.

• You can customize the label for the pitch to match your store’s style.

• Currently, the pitch is only available in the store’s default language.

• Including a pitch helps explain why the recommended products are relevant, encouraging customers to add them to their cart.

7. Automatically Rotating Recommendations

• If some recommendations are not converting, you can enable the automatic rotation feature.

• Selleasy AI will attempt to re-generate recommendations for products that are underperforming over time.

• This ensures your upsell offers remain fresh and effective without manual intervention.

• The “Automatically rotate when 50% or more recommendations are out-of-stock” option is enabled by default and cannot be disabled.

• When 50% or more of the products in a recommendation set go out of stock, Selleasy AI will automatically generate a new recommendation for that product with valid, in-stock products.

• If you need this option disabled for your store, please contact our customer support team.

8. Warming Up Process

9. Viewing Generated Recommendations

• The recommendations can be viewed via two places.
• By clicking the ‘View recommendations’ in Campaigns dashboard. This will display all recommendations generated so far in the store for all campaigns.



• By clicking the ‘View recommendations’ inside a respective campaign. This will display all recommendations generated so far only for that respective campaign.

• You can view the recommendations history in two types of views:
• Grid View: Quick visual overview of recommendations



• Table View: Detailed metrics for each recommendation

• You can view a recommendation in detail by clicking “View” on a particular recommendation in Grid view, or by clicking directly on the recommendation in Table view.

• Recommendation Statuses:
• Active: Recommendation is live and can be shown on the storefront
• Generating: AI is still creating the recommendation; click Refresh after a few minutes
• Rotating: Another recommendation is being generated to replace the current recommendation.
• Rotated: The particular recommendation has been rotated and no longer active.
• Archived: The recommendation has been archived and no longer active.
• Inactive: Campaign is in draft or scheduled status
• Campaign Deleted: Associated campaign has been removed
• Empty: No recommendations could be generated based on the current AI recommendation settings. Please update the settings and rotate this recommendation.

• Detailed View: Click on a recommendation to see:
• AI-generated pitch
• List of recommended products
• Recommendation status and creation time
• Performance metrics, including views, clicks, and conversions
• Rotate - Recommendation can be rotated, i.e, a new recommendation can be triggered to generate. This will consume AI credits.
• Hide - Recommendation can be archived, making it no longer active

10. Editing the Generated Recommendations

• Edit the AI pitch: If the automatically generated pitch isn't suitable, you can modify it to better match your preference.

• Edit the recommended products: You can remove products from the recommendation or add new products into it.

11. Changing or Customizing an Existing Campaign

• If you update campaign settings, such as editing the options below,
• Maximum number of products to show
• Recommended product type
• Instructions to AI

then upon saving, you will be prompted with two options:

• Apply Changes for New Generations:
• Only new recommendations generated after the update will reflect the changes. Existing recommendations will remain unchanged.

• Regenerate All Recommendations:
• All previous recommendations are discarded.
• New recommendations are generated through the warming-up process based on the updated settings.

12. Selleasy AI Credits and Pricing

• Fast model – consumes 1 AI credit per recommendation

• Thinking model – consumes 2 AI credits per recommendation

• Includes 50 AI credits for lifetime (50 recommendations)

• Once the credit limit is reached, it won’t be automatically recharged. To continue enjoying AI credits every month, please upgrade to a higher plan. The credits included in this plan are provided only for trying out the Selleasy AI feature.

• Includes 1000 AI credits/month (1000 recommendations per month)

• Custom pricing with typically 1000+ AI credits/month, where the credit limit can also be negotiated.

• Please contact our priority support for further details.