Skip to main content

[Organizing Production - GCW] Packaging Recommendations Configuration

Anastasiia Zbrozhek avatar
Written by Anastasiia Zbrozhek
Updated over a week ago

Packaging recommendations in GelatoConnect help partners ensure that every product is packed in the most suitable box based on product type, dimensions, and order quantity. By configuring this system, you gain full control over your packaging logic, reduce shipping surcharges, and improve operational consistency. This guide walks you through setting up your box inventory, creating matching rules, and using AI insights to refine your configurations over time.


Getting started with the box setup

Before creating packaging rules, you need to register all available box sizes in the system so they can be used for matching.

Upload your box inventory

  1. Set up the Packaging Product Model
    Create a model for your packaging material with essential attributes like:

    • Dimensions

    • Material type (e.g., Corrugated)

    • Weight capacity

  2. Create a customer-specific packaging product
    Use the product model to create a uniquely named product for each customer (e.g., “Outer Box SP”). This is what you’ll reference when configuring packaging logic.

  3. Define base attributes
    Assign fixed values such as shape (e.g., Rectangular), standard dimensions, and consistent labels for easier rule matching.

  4. Configure variant mapping
    Map external labels like “Poster Mailer” or “Medium Tube” to internal Gelato dimensions and attributes. You can support variations using aliases like “Med Tube” or “MT.”

  5. Save the configuration
    Once saved, these box products can be used in your packaging recommendation setup.


Automatically importing boxes from purchase orders

To minimize manual input, GelatoConnect can import packaging materials directly from recent GCP purchase orders:

  • Each time a purchase order with outer boxes arrives, the system checks if the boxes already exist in the product catalog.

  • If not, the system automatically creates them.

  • You should periodically review and remove unused boxes from your inventory.


Defining packaging recommendation rules

Packaging rules link product types or attributes to specific boxes. You can create different rule types depending on how you'd like packaging to be determined:

Types of rule logic

  • Fixed Box Mapping
    Example: Mugs → Mug Box

  • Quantity-Based Mapping
    Example: 1–6 posters → Tube; 6+ posters → Rectangle Box

  • Virtual Box Calculation
    Based on volume and box dimensions, using system logic

  • Advanced Attribute-Based Logic
    Example: Assign all flat products with 10x8 paper format to a specific box, regardless of quantity

You can also group multiple product models (e.g., flat cards and canvas prints) under one rule when their packaging needs align.


How to create a rule

  1. Go to the Product–Box Rules section

  2. Click Add Recommendation

  3. Select the desired box product

  4. Define product filters, including:

    • Product category or model

    • Attributes like paper format or size (e.g., 10x8 inches)

  5. Set the quantity range (e.g., 1–50 units)

  6. Click Save

Once saved, the rule appears in your list and clearly outlines:

  • Which box will be selected

  • Which products it applies to

  • The applicable size and quantity conditions

You can edit or delete rules anytime, and each change is logged in the Activity History tab for transparency.


AI-powered rule suggestions

GelatoConnect uses AI to identify when operators frequently override packaging recommendations. The system then suggests adjustments to improve future rule accuracy.

Here’s how it works:

  • The AI detects override patterns at the packaging station

  • It pre-populates a new rule form with suggested adjustments

  • You can review and update conditions (e.g., relax a size constraint)

  • Once confirmed, the new rule goes live for upcoming packages

Example:

If operators regularly choose a smaller box for flat 10x8 cards than what's configured, the AI will suggest a more suitable default to reflect this behavior.


Operator experience at the packaging station

When an order arrives at the packaging station, the system recommends the most appropriate box based on defined rules, AI suggestions, and carrier compliance requirements.

Operators will see:

  • The recommended box SKU

  • Box dimensions

  • Any cost or surcharge risk warnings

They can then:

  • Accept the suggested box and proceed

  • Override the suggestion by selecting another box from an approved list

If a different box is selected:

  • The system verifies if the override is acceptable

  • Warnings are issued if it could trigger shipping penalties

  • The choice is logged for AI feedback

If there’s a pre-generated shipping label, the system checks if the chosen box matches the label dimensions before proceeding.


Best practices and safeguards

  • Override monitoring: Manual overrides trigger AI review to keep recommendations up to date.

  • Compliance enforcement: Box mismatches are flagged early to avoid delays or repackaging.

  • Cost optimization: Recommendations help reduce dimensional surcharges and wasted packaging.


Troubleshooting

  • Box missing from dropdown:
    Check if the box is in the product catalog. If not, add it using the steps above.

  • Frequent AI alerts:
    Review logs to identify consistent override patterns and adjust your rules accordingly.


FAQ

What happens if no rule matches a product?
The system will fall back to manual box selection. You can review the product and create a new rule if needed.

Can I delete or edit a rule after it’s saved?
Yes, all packaging recommendations can be edited or removed at any time.

What if I want to override the system suggestion manually?
You can still choose a different box at the packaging station. This feedback is logged and used to refine future rule suggestions.

Does the AI agent update rules automatically?
No, it only suggests changes. You need to confirm and apply any updates manually.


📝 Not what you needed?

Help us improve this article, send us an email to [email protected] — please include the article title.

Did this answer your question?