Skip to main content

Product Tagging

An overview of how Product Tagging works in Creative Force, including: set up, configuration, building, and testing of Schemas & Attributes.

Written by Thomas Goblet
Updated over a week ago

Product Tagging is a powerful AI-driven feature that automatically extracts detailed product attributes from your studio images, transforming visual content into structured, revenue-driving data.

Add On: Product Tagging is part of an add-on package supported by the Workflow V2 engine. If you are interested in using it in your studio, please reach out to your CSM.


Tagging Overview

With Product Tagging enabled in your Studio, you can identify key details like material, fit, style, and more from production assets– eliminating manual tagging errors and dramatically reducing the time needed to enrich your product catalog.

Implement 80+ pre-defined attributes across categories like Consumer, Design, Physical, Style, and Usage. In addition, studios have the option to develop and add custom attributes specifcially tailored for your data.

With Cloud Automation, you can trigger Product Tagging automatically during Production, or trigger data updates manually as needed.

Attribute Data can be viewed, edited, exported & downloaded, or reused as metadata to enrich assets. Control is flexible and easily changed– ensuring that however Product Tagging works best for your studio, is exactly how it works, every time.

Tester

You can use the Product Tagging Tester to preview the types of attributes that AI can identify from product images.

To use the Tester, head to Studio Settings → Connectors → Creative Force AI → Product Tagging → then, click the 'Try It' button.

🛠️ Permissions: To use the tester, users need to be in a role with View or Edit permissions for Connectors.


In the tester, you can select from four Fashion Schemas (Tops, Bottoms, Dresses, and Purses) and two Label Schemas (Nutrition and Beauty).

The options within the tester for Attributes are common Attributes for the selected Category, and not exhaustive. All 80+ default attributes and custom attributes are supported in the full Product Tagging feature.


​Once a Schema is selected, you can select images to tag by uploading from your computer or selecting from assets in Creative Force. Then, just click 'Generate Attributes.'

Getting Started

Product Tagging is available for studios with the Product Tagging Add-On enabled for their studio. As a reminder, Product Tagging requires Workflow Engine V2 to work in production Workflows.

Enable Product Tagging

Before you can use Product Tagging, ensure the supported Connector is enabled for your Studio. Head to Studio Settings → Connectors → Creative Force AI:

In the pop-up menu, head to the 'Product Tagging' tab, and enable the sub-function.

Workspace Configuration

Head to Studio Settings → Workspaces → Hover over the intended Workspace → And select 'Edit.'

In the pop-up menu for the Workspace, scroll down to the 'Product Tagging' tab. Here you will find two configuration options to note.

Workspace Setting

Function

Schema Key

Select which product property will be used to identify attribute schemas from the available drop-down box.

Automated Tagging Updates

Toggle On/Off to allow regeneration of product tagging data if the schema trigger is updated during production.

Apply to Completed Workflows

Becomes available when Automatic Tagging Updates is enabled.

Toggle On/Off to allow regeneration of data even if the Workflow has already been completed.

Schema Key Selection

Schema Keys work in tandem with Attribute Schemas and Triggering to support automatic updating of attribute data based on conditions.

The 'Category' property is the default Schema Key, but you can also select the property of your choice. After selecting the property, you will need to define which property values should trigger the Attribute Schema.

For now, save your Workspace settings, and we will cover the Attribute Schema & Trigger process in the next section.

Attribute Schemas

With your Workspace settings configured and your Schema Key selected, you'll need to set up an Attribute Schema within your Studio Settings

The Schema acts as a blueprint that defines which Attributes to extract and their acceptable values for specific Product Categories. The Product Category will be used to map the Schema and Products together.

Navigate to Studio Settings → Data Management → Attribute Schemas

Edit Schema

To edit an existing one, simply click 'Edit' when hovering over the schema of your choice.

Create New Schema

To create a new schema, you can click on 'Add Schema' in the top right corner of your screen.

Schema Settings

In the initial tab of the Attribute Schema menu, you will find all of the available configuration options for the Schema and how it should behave.

Schema Option

Function

Schema Name

Give the Schema configuration a unique name for internal identification.

Workspace

Choose the Workspace that this configuration should be related to.

Enabled

Toggle On/Off, controls whether the Schema is currently enabled and available for use in Productions.

Triggers

Add specific values that should act as triggers to re-generate attribute values during production.

Triggers Note

The Schema Key for the Trigger will be auto-inherited from the selected 'Schema Key' in the Workspace settings.

The listed values for triggers that can be added are dependent on the products that have been imported to the Workspace.

If there are no imported products with values of the property you have selected as the Schema Key, there will be no values to add as triggers.

Attributes Tab

In the Attributes Tab, you will be able to see all of the Attributes that are currently configured for this Schema, their Category as represented by the Category icon, and their related return values.

If you are creating a brand new Attribute Schema, this tab will be empty. Before attributes appear here, they will need to be added via the Schema Builder.


Schema Builder

It’s now time to start building your Schema. To get started, select the 'Schema Builder' button on the bottom left-hand of the Attribute Schema menu.

Attributes List

On the left-hand side in the resulting pop-up menu, you’ll find a list of attributes. By default, there are around 80 attributes, but custom attributes can also be created.

The existing attributes will be sorted into five main categories:

  • Consumer

  • Design

  • Physical

  • Style

  • Usage

Use the search bar to look for a specific attribute across all categories:

Selecting Attributes

If you want to add an attribute, simply tick its box, and it will automatically be added to the schema builder visible on the right-hand side.

In this example, we have selected the attributes Fit Type and Height Recommendation. In the builder, you can see the attributes in purple, with a list of the potential return values in blue.

When you have added all the desired Attributes, select 'Apply,' and the Attributes will be generated into the Attributes tab back in the main menu.

Edit Attributes

To edit an existing attribute, you can click on the logo near the attribute, then click 'Edit.'

For more information on the available Attribute Fields that can be edited and their individual functions, see the table in the next section.


Create New Attribute

You also have the option to add new attributes, which will then become available for your current Schema, or any other new or existing Schema in the system.

In the table below, we will review the available configuration fields when creating or editing an Attribute.

Attribute Field

Function

Attribute Name

Name the attribute for internal identification within the system.

Category

Select a pre-defined Category from the drop-down list. This defines how the attribute will be organized.

View Categories

  • Consumer

  • Design

  • Physical

  • Style

  • Usage

Internal Name

A unique identifier automatically defined by the system. This property can not be edited or changed.

Description

The description can be used as a prompt for the AI to understand what it should be looking for when performing tagging.

eg, "Summer Dresses are light colors and fall above the knee."

You can also include negative prompts.

eg, "Summer Dresses are not two-piece matching skirt and top sets."

Returns

Define whether the system must return a single specific value or if it can return any range of applicable values.

  • Single Value: Return 1 value.

  • List of Values: Return an array of values.

Potential Return Values

Add all the expected values that can be potentially returned for this attribute.

  • Leave empty to allow any return value

  • Define a single value on each line

🗣️ Tip: Leaving the 'Potential Return Value' field blank allows the AI model relative freedom in defining the attribute values.

Testing Your Schema

The Schema Builder includes a built-in tester that lets you validate your configuration before adding it to a live Workflow.

To access it, click on the 'Tester' tab along the top of the Schema Builder Menu:

Choose a Product Code within Creative Force by searching in the box to the top left.

🗣️ Tip: When searching for Products, they must be part of the same Workspace as the Schema you are currently building.

If you have multiple images for the same Product, you can also select a specific one that should be used for the test.

After confirming the product and assets, click the 'Generate Attributes' button at the lower right of the screen.

The AI will begin to analyze and return relevant properties using the configuration for the Schema as you have built it.

The analysis typically takes 1-2 minutes, and you can refine your Attribute configurations based on the output results.

If an attribute can't be identified from the image, it will be left blank rather than guessed incorrectly.


Add to Workflows

In your Workflow builder, navigate to the step where you want to add Product Tagging, and 'Add' a new Cloud Automation function.

Select a relevant connector (such as Creative Force AI) → then select 'Product Tagging' as the sub-function to be executed.


The most common placement is post-Final Selection, to ensure only approved images are processed. However, it can be positioned anywhere prior to Asset Delivery that best serves your use case.

While it's possible to have multiple subsequent steps of Product Tagging within the same Workflow, the last iteration of the step will simply override previous sessions' data. As such, it is not productive unless for a very specific use case.


When the Product reaches the specified Workflow step while moving through production, the assets will be checked by AI, and attributes will be updated based on the Schema settings you have selected.

​Applying Product Tagging

In addition to the automated Workflow tagging that can be performed in conjunction with Cloud Automation, Product Tagging can be applied manually either to an individual product or a group of products in bulk.

Individual Tagging

To apply Product Tagging to a specific Product outside of production, head to the product overview → Under 'Product Tagging' select 'Add,' → then, 'Generate Attributes.'

Bulk Tagging

You can also use Product Tagging to update Products in bulk outside of Workflows. This is especially useful for Products that completed their Productions before Product Tagging was enabled.

To use Bulk Tagging, head to the Product Tab, filter for, and multi-select the desired Products. Use the 'Actions' button → then select 'Product Tagging' from the actions menu.

​In the pop-up menu, you will have a number of options to define how you would like to apply your bulk tagging.

Option

Function

Schema

Select an option from the drop-down menu.

  • Choose a pre-configured Workspace-Specific Schema for the relevant Workspaces.

  • or 'Based on Schema Trigger.'

→ Based on Schema Trigger

The AI will loop back to the Schema Key and Trigger settings you previously defined in the Workspace Schema Settings.

Tag Specific Attributes

A new pop-up menu will appear, allowing you to select specific attributes to be tagged or retagged.

Tag All Attributes

The AI will attempt to find and return values for all Attributes configured within the Schema.

Once you have triggered the bulk tagging session, the automation will start to work in the background. You are free to move into other screens and complete other tasks.

To check the progress of your bulk tagging, select the bell icon towards the top of your screen to review the notification center. The system will alert you when the task is completed, or if any errors were encountered when attempting to apply the data.


Managing Tagged Data

Once Products have had attribute data generated, either manually or by automation, there are a number of options to further manage and review that data within Creative Force.

Review Attribute Data

For Products that have successfully had Product Tagging data applied to them, you can access the results from the Product slide-in menu.

To review data, navigate to the desired Product → select it to open the slide-in menu → then, select 'View Attributes.'


From the resulting pop-up menu, you can check each Product Attribute for accuracy and edit values manually as needed.


Change History

A version history of any changes applied to product attributes can be seen and reviewed by clicking on 'Change History' in the top right.

On the right-hand side, a new menu column will appear. If changes have been applied to the attribute data, they will be shown here. Select between the different versions to review the data as it was.

The selected version is shown in the furthest bottom left corner of the menu, along with an associated date/time.


Download Attribute Values

Further down, you will also have the option to download a list of the current Attributes and their values. Just select 'Download' towards the bottom-right corner of the menu, and choose your preferred file type for the export.


​If there are multiple 'versions' associated with changes to the data over a period of time, the values downloaded will be the data relevant to the version you have currently selected.

Product Tagging Tokens

Attributes generated by Product Tagging can be added to Copywriting content or as Metadata to assets.

To use Product Tagging tokens in Copywriting, you can add them as Merge Fields in your Copywriting Style Guides.


To use Product Tagging tokens as metadata, you can add them as Image Metadata in Presets.

Attribute Column Display

Attributes generated by Product Tagging can be viewed within the Product Tab.

To view Attributes in the Product List, head to the Column Display Settingsand select to enable them under the 'Product Tagging' section.

Notes & Call Outs

The following section contains important call-outs or notes related to Product Tagging. Please review these notes carefully, as they are relevant to frequent points of confusion or roadblocks.

Data Priority

⚠️ Important: AI uses a combination of analyzed product images and product properties while tagging products.

If a product property is available for an Attribute needed by the Schema, the product property will be used, rather than what is seen by the AI model.

Example: If the value of the CF Product Property for 'Color' is currently 'Red,' and there is an attribute for 'Color' in the Attribute Schema:

  • The AI will return 'Red' as the Color, even if the image features a product that is actually blue.

Supported Labels

⚠️ Important: Product Tagging is specifically designed to support apparel, nutrition, and beauty labels. ​

While labels of different types may technically return results, they are not tested for or guaranteed to any specific level of accuracy.

Supported File Types

While the AI models that power Product Tagging only accept specific file types and sizes, these limitations should not impact you. If images for products using Product Tagging don't meet the AI model specs, we will use the full-resolution preview JPEG that is created for all assets uploaded to Creative Force instead.

This means you can use Product Tagging with any file format in your Production Workflow without worrying about compatibility issues. Creative Force automatically handles the necessary conversion completely behind the scenes.

🗣️ Tip: For the input image, Ghost Mannequin images work particularly well for recognition with the current models.


Bulk Attribute CSV Export

Currently, Product Tagging Attribute Data can be exported & downloaded directly on a product-by-product basis. However, specific attribute data can be seen and downloaded in bulk using the Column Display settings in the Product Tab:

  1. Filter for the range of Products you would like to export attribute data for.

  2. In 'Column Display Settings,' enable the desired attributes to export.

  3. Multi-select the range of products.

  4. Select the 'Actions' button

  5. Select 'Export CSV.'

Did this answer your question?