

Credits per action
Every action shows its price in the modal footer before you start. These are the rates the price is built from:
Products that need no work, such as research that is still fresh or fields that are already filled, cost less than the ceiling. Credits for copy that was not saved are refunded.
Pick an action
The modal opens with a Recommended starting path: products selected, then Web research, then Enrich product content. Research adds the facts AI writes from, so running it first gives richer copy. Below it, type in Search actions… to filter the cards, or pick from the shelves:
Translation and Linguistic Review only appear when your project has target languages. Web research disappears when the feature is switched off for your workspace. When you select up to 20 products one by one, a card shows Recently completed · N/N for products that got a saved update from that action in the past 30 days. You can still run it again.
Run a single action
1
Select products
Tick products one by one, or filter and select all matching products. FAQs and HS codes only run on products you select on the page, not on all matching products.
2
Click Enrich with AI
The modal opens on Step 1 · Choose an action.
3
Choose the action
Content actions go to Step 2 · Choose a prompt first: pick one of your prompts or one curated by WISEPIM. Your project default is preselected and listed first.
4
Configure and start
Set the options for the action, check the cost in the footer, and click the start button. Its label names the action, for example Start Enrichment, Start Web Research or Generate FAQs.
If the start button is greyed out, the reason sits right next to it: “Choose an enrichment action.”, “Choose at least one target language.”, “Choose at least one field to translate.”, “Choose at least one action.”, “Choose a prompt for this enrichment.”, “Pick at least one FAQ type”, or “HS code classification runs on selected products. Select products on this page instead of all matching products.” While the price is being checked it reads “Checking eligible products and credits…”.
Run a workflow: Enrich product content or Fill empty fields


Under Enrichment actions, tick what the run should do. The cards are grouped as Data Foundation, Product content, SEO & discovery and Languages & quality. Titles, Descriptions, Short Descriptions and SEO Title and Meta Description start ticked. Attributes is opt-in.
- Each content card shows its prompt. A ticked Titles, Descriptions, Short Descriptions, SEO or Attributes card carries a Prompt picker, set to Project default (name). Pick another prompt for this run only. A field with no saved prompt says “No saved prompt for this field yet. Create one in Prompt Lab.”
- The order is fixed. The header reads “N actions selected · Ordered automatically. Each selected step finishes before the next begins.” Whatever you tick runs as: web research, categorization, attributes, titles, description, search intent, short description, SEO, internal links, image alt text, FAQs, translation, quality review.
- Some actions add options. Web research asks for the identifier and Include images from research. FAQs has its own options. Translation needs Target languages and Fields to translate, and can Use Pro translation (5 credits per language and product). Review product quality can Include images in the quality review.
- Search Intent and Internal Linking rewrite descriptions, so they only run in Replace selected content.
Check the cost before you start
The price is calculated by WISEPIM for the exact run you are about to start, while the modal is open. The footer shows Up to N credits · N available.- No products need the action. Start stops with “No products are eligible for this action.”
- Your balance is too low. Start stops with “This run needs N credits and you have N.”
- A large run (500 credits or more). A confirmation asks first: Start this run?, or Review workflow estimate for a workflow. Click Start run to go ahead.
- Something changed after the price was shown. If the project, selection or settings changed, you see “The project, selection or settings changed. Check this run again.” The run never starts at a higher price than the one you saw.
FAQs and HS codes have no price check yet. Their footer says “The cost is checked when the run starts.”, and the balance is checked at launch.
Content enrichment
The content actions follow the same flow: select products, pick the action, choose a prompt, start. The AI combines the prompt with each product’s data, its web research and your Knowledge Library.

Whether AI changes are saved right away or wait for your approval is set on Settings > AI review. It is off by default, so AI work reaches your catalog without extra steps.
When credits run out mid-run
The modal does not start a run your balance cannot cover. If credits run out while a run is going, for example because another run spent them, the task stops with a clear message instead of failing silently. In a workflow the stage that ran out says “Not enough credits to finish this stage.” and later stages wait. Once you have credits again, open the task in the Process Tracker and choose Resume. Resume picks up only the products that were never processed, at no extra cost.Resume finishes the remainder at no extra cost. Restart runs the whole selection again and is charged as a new run.
AI Categorization


Internal Linking


Search Intent


Image Alt Text


Translation


Web research


Linguistic Review and Review product quality
These two start from the modal but open their own workspaces:- Linguistic Review scores your translations and offers fixes for inconsistencies. It needs at least one target language in Translation Settings.
- AI Quality Review reads each product like a demanding customer, scores it and hands you a ranked list of fixes.
Single product enrichment
Every enrichable field on the product page has an Enrich button. It runs the enrichment on that one field of that one product, with the same prompt choice. Enrichable fields are the name, description, short description, SEO title and meta description, attributes and categories. When a run saves nothing, the product page tells you why, for example “AI enrichment didn’t change anything” with “Existing content was kept: there was no empty field to fill.”Using Content Logic with enrichment
Content Logic variables in your prompts give the AI per-product context. A prompt like:Automating enrichment
Enrichment actions are available as automation actions. Trigger them on a schedule, after imports, when quality scores drop, or when products change.Tips
- Start with 5 to 10 products. Test a prompt on a small batch and read the output before you run it on the whole catalog.
- Research first. Products with web research get richer, more accurate copy.
- Fill your Knowledge Library. The more the AI knows about your brand and audience, the more consistent the output. See Knowledge Library.
- Use specific prompts. A prompt for “B2B industrial fastener descriptions” beats a generic one. Build them in the Prompt Library.
Act on what you find
A workflow stopped at one stage
A workflow stopped at one stage
Open the run in the Process Tracker. The parent task names the stage, its counts and the cause. Earlier results are kept. Fix the cause, for example add credits or review pending category suggestions, then resume. Outcome: the rest of the workflow runs without redoing finished stages.
A content field you edited by hand was rewritten
A content field you edited by hand was rewritten
Replace selected content always replaces the enriched name, description and short description. Use Fill empty fields when you want to keep filled copy. Outcome: your hand-written copy stays.
Related
Prompt Library
Build and organize the prompts that power content enrichment.
Prompt Lab
Test and refine prompts on real products before running them in bulk.
Product FAQs
Generate buyer questions and answers for your products.
Knowledge Library
Give the AI your brand voice, audience, and product context.
Enrichment Impact
Compare how products sold before and after enrichment.
Teach WISEPIM your house style
Set up brand voice once so every AI-written description sounds like you.

