Enrichment uses AI to fill in your product data, images, and content automatically. This guide covers how to start an enrichment and understand the queue.
The three types of enrichment
Trustana offers three enrichment types. Each targets a different kind of product content.
Data Enrichment
Data enrichment fills in product attributes such as descriptions, specifications, ingredients, and other category-specific fields.
Trustana works from your existing data first — your imported product data and product images are analysed to extract visible information. To extend this with the web, turn on Enable AI to source data externally in the confirmation modal. With external sourcing enabled, Trustana also browses the web for additional context such as manufacturer specs, descriptions, and images.
Tip: Add a Product Model to your products to dramatically improve web-sourcing accuracy, especially for branded goods.
When Enable AI to source data externally is on, you can fine-tune where Trustana looks:
Select countries — restrict the search to specific countries
Preferred sources — prioritise up to 10 URLs Trustana should favour
Excluded websites — block up to 10 URLs from the results
Image Enrichment
Image enrichment finds and assigns product images from the web — based on product name, brand, and identifying details. It is designed for two scenarios in particular:
Offline-to-online catalogues — suppliers whose products currently have no images at all
Better classification — images are one of the strongest signals our AI uses to pick the right category, so getting an image attached makes the rest of enrichment more accurate
A globally searchable Product Model dramatically improves match quality. Variant-specific image search (different image per colour, flavour, etc.) is supported but requires Customer Success to enable it for your workspace — please reach out if you need it.
PDF / File Data Extraction
PDF enrichment extracts factual product data from files you upload — such as industrial catalogues, price lists, and marketing brochures. It matches products in the file to your catalogue and populates the relevant attributes.
When you start a PDF enrichment, you can turn on Include Additional Context to send surrounding pages to the AI alongside the matched page — useful for single-product files where a product's data spans multiple pages. Leave it off for multi-product tables and catalogues where each product fits on one page.
Eligibility
Before starting an enrichment, Trustana checks that each product has enough information to work with. The check runs at submission and again when the product is processed — so if data changes during a long queue wait, the check always reflects the current state.
Enrichment type | Required — parents | Required — variants | Recommended |
Data Enrichment | Product name, Brand, and at least one of: a product image, a Product Model, or a category already set | Same as parent | Product Model — unlocks External Data Sourcing for higher accuracy |
Image Enrichment | Product name, Brand | One variant-specific attribute filled (e.g. Color, Flavor) — or contact Customer Success to enable variant-specific image search | Globally searchable Product Model — significantly improves match quality |
PDF / File Data Extraction | Everything required for Data Enrichment, plus a Product Model or Supplier SKU and a matching record in an uploaded file | Same as parent | Clean, well-structured source PDFs |
If a product is missing required fields, you will see a warning banner — "This product cannot be enriched." — when you try to enrich it. Click View product in the banner to filter your product table and see all products with missing information.
You can also audit your catalogue at any time — independently of any enrichment task — by applying the Missing required information filter in the product table.
Note: If you select a mix of eligible and ineligible products, the task is still created. Ineligible products are automatically removed from the task — they keep their previous status and are not marked as failed.
Note: Products that are already part of another active enrichment task are also automatically excluded. Wait for the existing task to finish (or cancel it) before re-enriching the same products.
Choose what to enrich — Applicable or Specific Attributes
When you start a data enrichment, choose which attributes to target:
Applicable attributes — the system automatically enriches all attributes that apply to each product based on its category. This is the default.
Specific attributes — you pick exactly which attributes to enrich, up to 10 per task. Use this when you want to update a targeted set of fields without touching everything else.
When you select Specific attributes, the attribute picker opens automatically. Search for and select the attributes you want to enrich, then click Confirm.
Once attributes are selected, the button changes from Select attributes to Review Selection so you can always see and adjust your choices before submitting.
Note: Re-categorisation only happens in Applicable attributes mode. Specific attributes always keeps the product in its current category.
Re-enriching existing data
By default, enrichment only fills in empty or rejected fields. Turn on Include Enriched Attributes to also re-enrich fields that AI has previously filled. This toggle works in both Applicable attributes and Specific attributes modes.
When Include Enriched Attributes is on and you are in Applicable attributes mode, two additional options appear:
Re-enrich on category change only — the AI re-classifies the product first. If the category stays the same, previously enriched attributes are skipped. If the category changes, those attributes are re-enriched against the new category. Tip: Skip re-enrichment if the category does not change to avoid wasting credits.
Remove attributes that no longer apply after re-categorization — when a product moves to a new category, attributes from the old category are automatically removed.
Tip: Turn on Re-enrich on category change only when running bulk re-enrichments. It saves credits on products whose category has not changed.
Note: In Specific attributes mode, the re-categorisation options above do not appear — re-categorisation only happens in Applicable attributes mode.
AI Rules & Context Graph
AI Rules let you add brand- or category-specific instructions to the enrichment prompt — for example, a tone of voice guide, a translation glossary, or category-specific writing rules. These are applied automatically whenever products matching that brand or category are enriched.
Start an enrichment
Go to All Products from the top-left menu.
Select the products you want to enrich using the checkboxes. You can also use Select All to select your full filtered set.
Click Enrich Data (or Enrich Images for image enrichment) in the bottom-right toolbar.
In the confirmation modal, choose your enrichment scope (Applicable attributes or Specific attributes) and configure any re-enrichment options.
Click Start Enrichment to submit.
Tip: You can also trigger enrichment for a single product directly from its product detail page — open any product and use the Enrich action in the top-right.
Note: If any selected products are not eligible, they are automatically removed before the task starts. The task proceeds with the eligible products only.
The queue
After you submit, the task enters the queue. Trustana processes enrichments continuously — you can submit on a Monday and the system will work through the weekend. Tasks wait in the queue for up to 96 hours before timing out.
Trustana runs one task at a time. Each task supports up to 1,000 products in total — parents and variants combined. For large Select All operations across thousands of products, the system automatically batches the selection into multiple tasks behind the scenes. There is nothing extra to do — it works as a single click.
Task statuses, Cancel & Retry
Go to Enrichment Tasks from the top-left menu to monitor your tasks. Use the Needs Attention quick filter to surface tasks that require action.
Pending | Task created — eligibility check running |
Queued | Eligible products confirmed — waiting to start. Can be cancelled. |
In Progress | Actively processing |
Finished | All products enriched successfully |
Finished with Issues | Completed but some products had errors — use Retry |
Failed / Timeout | Task did not complete — use Retry |
Cancelled | Cancelled by the user before any product succeeded |
Aborted | No eligible products found — nothing was processed |
Cancel is available while a task is Queued, or In Progress before any product has been successfully enriched. Once at least one product succeeds, use Retry instead.
Note: The View Products link inside a task shows only the products that passed eligibility and are part of that task — not the full original selection. To see products that were excluded, use the Missing required information filter in the product table.
What's next