Filling metafields on your products
Two custom jobs write product metafields. Fill one metafield on your products fills a single metafield in a review table, 25 products per page. Fill several metafields on your products fills two or more in one run, each with its own settings. Both use the metafield definitions already in your store, and both need a paid plan.
How to fill one metafield
- Open Build a custom job and click Next on Fill one metafield on your products.
- Under Which metafield do you want to set?, tick one metafield and click Next →.
- Choose the products in the filter bar and click Apply.
- Set Mode. With Manual, set one value and click Fill all. With AI, choose Sources, add AI guidance if you want, and click Fill & review.
- Check the rows. Edit a value in its row, or click Reset to drop it.
- Click Save this page. The toast offers View to open the run's report.
To fill every matching product at once, open More actions and choose Bulk fill & save (AI) or Save all (Manual). You confirm the count and cost, then the job runs in the background.
How to fill several metafields
- Click Next on Fill several metafields on your products, tick two or more under Which metafields do you want to fill?, and click Next →.
- Pick a metafield in the list on the left and set its Mode.
- Click Fill samples (AI) or Preview value (Manual). Up to five products show under Product Samples.
- Click Approve. Repeat for each metafield.
- When every metafield is approved, click Save all and confirm.
Good to know
- The list shows your own product metafields. Shopify's standard category metafields are not listed (use Set category attributes), and JSON, rich text, and customer or company references are not offered.
- Reference metafields, such as product, variant, collection, file, page, article and metaobject, open a picker instead of a text box.
- Samples are free and are never saved. Changing a setting clears that metafield's approval.
- Save all for several metafields starts one job per metafield, so three metafields appear as three runs in Recent activity.
- AI credits are charged when an AI-written value is saved. Manual values cost no credits.
- Whole-catalogue runs use the filter you applied. While a filter change is waiting, the menu item reads apply your filters first.
- Filtering by a metafield for the first time can ask you to Enable filtering.
- The run's report links back to the page it came from, for example "Back to Set Material".