Skip to main content

Understanding trigger types

Triggers determine when products enter an AI Workflow.

Every workflow starts with a trigger. The trigger defines:

  • which products should be processed

  • when the workflow should run

  • how products move into the automation pipeline

Without a trigger, a workflow cannot execute actions.


How triggers work

Triggers use rule based logic to monitor and select products.

When a product matches the configured conditions, the workflow can start processing it.

Example trigger rules:

  • Description is empty

  • Category equals Running Shoes

  • Status is enabled

  • Brand equals Nike

Only products matching the configured rules enter the workflow.


Available trigger types

AI Workflows currently support the following trigger types:

  • One-time triggers

  • Continuous triggers

Additional trigger types, such as import based triggers, may be added in the future.


One-time triggers

A one-time trigger processes all currently matching products once.

After execution is completed, the workflow stops automatically unless manually restarted.

This is useful for:

  • bulk enrichment projects

  • initial catalog cleanup

  • large translation runs

  • supplier data improvements

  • one time migrations


Example

Trigger rules:

  • Description is empty

  • Status is enabled

When the workflow starts:

  • all currently matching products are processed once

  • newly added products are not processed automatically afterward


Continuous triggers

A continuous trigger continuously monitors your catalog.

Whenever a new product matches the configured rules, the workflow automatically processes it.

This creates an ongoing automation pipeline.


Example

Trigger rules:

  • Shopping description is empty

  • Category equals Electronics

Whenever a new matching product appears:

  • the workflow starts automatically

  • actions execute automatically

  • no manual intervention is required


Continuous workflows are useful for

Continuous workflows are ideal for:

  • automatic enrichment pipelines

  • ongoing translations

  • catalog quality monitoring

  • automated validation

  • recurring synchronization processes

They help keep product data continuously updated over time.


Rule based targeting

Triggers use the rule builder to define which products should enter the workflow.

This allows highly targeted automation.

Examples:

  • Only enrich products from a specific supplier

  • Only translate products missing German content

  • Only process products missing attributes

  • Only validate products ready for publishing

This prevents unnecessary AI processing and helps control workflow scope.


Trigger entities

Triggers operate on a selected entity type.

Currently supported:

  • Product

Future entity types may include:

  • Categories

  • Pages

  • Blocks

  • Other catalog objects


Source store selection

Triggers can also target a specific store scope.

Examples:

  • Global

  • Store specific catalogs

  • Regional store views

This allows workflows to operate only on relevant store data.


Product count preview

While configuring a trigger, Elovate displays how many products currently match the configured rules.

This helps you:

  • validate your conditions

  • avoid accidental large executions

  • estimate workflow size

  • control AI processing volume

Example:
A trigger may show:

  • 56 matching products

  • 12,000 matching products

  • 3 matching products

before the workflow is created.


Editing triggers

Triggers can be updated after workflow creation.

You can:

  • change rules

  • modify conditions

  • update workflow scope

  • pause workflows

  • reactivate workflows

This allows workflows to evolve with your catalog strategy.


Trigger best practices

Start with small scopes

When testing new workflows:

  • use limited product sets

  • validate prompt quality

  • review outputs carefully

before scaling to larger catalogs.


Use precise conditions

Avoid overly broad workflows.

Good example:

  • Category equals Running Shoes

  • Brand equals Nike

  • Description is empty

Less effective example:

  • Status is enabled

Precise targeting improves output quality and reduces unnecessary processing.


Separate workflows by purpose

Instead of one large workflow handling everything, create focused workflows.

Example:

  • Translation workflow

  • Attribute extraction workflow

  • SEO enrichment workflow

This improves maintainability and debugging.


Future trigger types

Additional trigger types may be added over time.

Planned examples may include:

  • Import based triggers

  • Synchronization triggers

  • Scheduled triggers

  • Event based automation

Example future scenario:
"When an import finishes, automatically enrich and translate all imported products."


Example trigger flow

Example:
A webshop wants to enrich eco products automatically.

Trigger:

  • Category contains Eco

  • Status is enabled

Actions:

  1. Extract sustainability attributes

  2. Generate SEO descriptions

  3. Translate content

  4. Synchronize approved output

Whenever products match the trigger rules, the workflow begins automatically.

Did this answer your question?