A catalog is how you get products into Renable. Typically you'd set up one catalog per store or market — for example, one for your main website and another for a regional storefront.
A catalog belongs to exactly one project, decided when you create it, and that assignment is permanent — there's no way to move a catalog to a different project later. If you need the same products available in another project, you create a separate catalog and import the feed again; the two won't share import history or enrichment state. See Managing catalogs in a project for how catalogs are added to and removed from a project.
Each catalog controls:
- Where its products come from (a product feed)
- How that feed is shaped and cleaned before it reaches your ads — mapping, rules, and joined data
- Optional reviews and top-seller statistics
- Which AI enrichment features run on its products
- Its locale
- AI prompts that apply to all of this catalog's products, in any project
Importing products#
Every catalog reads its products from one product feed, and the Product feed source dialog is where you say where that feed lives. Four kinds of source sit side by side at the top of it:
- Feed address — a direct link to your feed file, hosted by your e-commerce platform or a feed-management tool. This is what most catalogs use, and Renable works out whether the file is XML, JSON, CSV or TSV by reading it. The field is labelled Feed URL.
- Google Sheet — a sheet published to the web, pasted into the Google Sheet link field. See Connecting a Google Sheet.
- Upload a file — a file you upload here and replace yourself when it changes, for products that have no feed URL to point at. See Uploading a feed file.
- Ad platform catalog — read the catalog you already keep at Meta or Google instead of a file. See Importing a Meta catalog and Connecting Google Merchant Center. If you would rather point at a public feed URL than connect an account, see Finding your feed in the Google Merchant Center.
Whichever you choose, the dialog reads the source as you type and shows a preview of the columns and the first rows it found — or the reason it could not read them — so a wrong address or an unexpected format is visible before you save.
The first time you connect a feed, "Save & generate mapping" reads it all the way through before suggesting how its columns line up with catalog fields, so a column only some of your products carry is picked up too — an uploaded file the same as a feed URL. While that read runs the button reads "Analyzing feed…", with the number of rows read so far once there is one, and waits for it to finish. If the read can't be completed the button goes back to its ordinary label and the dialog tells you the mapping will be suggested from the preview rows instead, pointing you at the Mapping tab's refresh to read the whole feed later — so a feed Renable struggles to read end to end never leaves you stuck. See Mapping your feed fields for what you get and how to adjust it.
The preview rows are always fetched for the preview itself, so they show you what your feed is serving right now. The feed's column list beside them can be older — it is read from the copy Renable already had — and when the two come from different reads the dialog says so under the counts and asks you to refresh the columns on the Mapping tab after saving.
If you need a fresh import right away, the Import now button in the catalog's health strip runs one on demand — against your last published configuration, not your draft, so publish first when you want the edits you have just made to be applied. Feeds are checked for changes every hour. By default Renable picks the time within the hour automatically; the Import time setting in the Product feed source dialog lets you pin it to a specific minute past the hour (in 5-minute steps) — useful when your shop regenerates its feed at a known time, so each import picks up the freshest file. See Questions about catalogs for what happens when a feed briefly fails, and how product limits affect what gets imported.
Shaping and fixing your feed#
Importing is only the first half of a catalog's job. Once your products are in, the catalog workspace is where you fix and shape that feed before it reaches your ads — rename columns, clean up bad values, drop products you can't advertise, join in extra data, and line everything up with the fields ad networks expect. Every change stays in a draft, and nothing reprocesses or reaches a channel until you Publish. See Transforming your feed for how the workspace and its draft-and-publish model fit together.
The work splits across a handful of tools:
- Mapping your feed fields — line your feed's columns up with the catalog fields channels expect, so your title column feeds the Title field, your price column feeds Price, and so on.
- Fixing your feed data with rules — conditions plus actions that clean and reshape values: set or remove a field, exclude a product, or split it into variants.
- Joining another feed — pull in a second feed for stock levels, supplier data, or extra fields, matched to your products with a lookup.
- Mapping tables — convert many source values to correct ones in bulk, such as mapping your own category names onto the categories an ad network expects.
- Working with rulesets — when several catalogs need the same cleanup, keep those rules in one place and apply them everywhere.
- Splitting variants into products — turn one feed row into separate products, one per size, color, or other variant.
- Feed runs and warnings — see what each import did, read any warnings, and trace how a single product's values were produced, step by step.
You don't have to work through these by hand. Describe what you want to Aimée in plain language — "exclude everything that's out of stock," "map my category column to Google's categories," or "join this stock feed" — and she can set the feed source, edit the rules and mapping, join another feed, and publish once you approve.
A note on barcodes#
GTIN/EAN/UPC values from your feed are only used for two things: filtering product sets and matching incoming customer reviews to the right product. They don't merge or unify products across catalogs — matching for content purposes is by SKU, scoped to a single project. See How Renable works for how products from different catalogs in a project are deduplicated in the UI.
Review feed (optional)#
A catalog can also import customer reviews, which Aimée uses to pull quotable fragments and ratings into ad copy. See Connecting a review feed for supported sources and setup.
GA4 top-seller ranking (optional)#
You can connect a Google Analytics 4 property (Connect to GA4, a separate button from the product-feed connection) so Renable knows which of your products actually sell best. This doesn't import any products — it only supplies ranking data, used to prioritize which products get processed first when you're near your plan's limits, and to help you build "best sellers" product sets.
Don't confuse this with Connecting Google Merchant Center — that's a completely separate Google connection that does import products.
AI enrichment features#
The catalog's Features section controls which AI processing runs on this catalog's products:
- Automatically create product images with transparent background — turns on the background-removal pipeline (see below).
- Automatically categorize images as product, lifestyle, staged, .. — classifies each image's content type.
- Automatically extract ad copy from product descriptions — pulls a short fragment straight out of existing product descriptions, useful when your descriptions are already well-written. (AI-written ad copy is configured with prompts on the project's Creative tab instead.)
- Automatically predict Google product categories when they are missing in the feed — fills in
google_product_category when your feed doesn't supply it. Turning it on fills the category for every product in the catalog that is missing one; turning it off removes the predicted values again. A category your feed supplies is never touched either way.
- Use reviews from all products with the same Group ID — pools reviews across variants (color, size, etc.) instead of only using reviews written on the exact SKU.
Content types#
Several places in Renable — including the background-remover pipeline below — let you scope something to specific content types. The full list: Information, Product, Lifestyle, Staged, Feature, Review, Award, Ad copy, Headline, Reason to buy, Voice script, Product name.
For image work specifically, the categories you'll use most are:
- Product — the product on its own, usually against a neutral or white background.
- Lifestyle — the product shown in a natural, real-world setting.
- Staged — the product shown in a studio setting, on a model, or otherwise professionally staged.
- Feature — a close-up of a specific product detail.
- Information — badges, certifications, infographics, and other informational graphics.
- Award — an expert-review badge or award seal, e.g. a "Best in test" logo.
- Review — images sourced from a customer review.
The remaining categories (Ad copy, Headline, Reason to buy, Voice script, Product name) tag written content rather than images, and mainly come up when reviewing or scoping AI-generated copy — see Reviewing and approving content.
The background-remover pipeline#
Turning on transparent backgrounds doesn't just pick a single AI model — it lets you build an ordered pipeline of steps, each with its own settings. It lives on the catalog's Settings tab, in the Image processing pipeline section. For every image, Renable tries the first step; if the result doesn't meet that step's quality bar, it falls through to the next step in the list. If no step scores high enough, the best result of them all is kept and the image is flagged for review.
Each step has:
- Background remover — which model to use: Make transparent (v1), (v2), (v3), or No transparency (skip background removal for this step; the image is still cropped if the step's crop switch is on).
- Use for content types — which categories this step applies to (see above). Leave it empty and the step applies to product images only.
- Automatically crop away image backgrounds as much as possible — trims empty space around the product after background removal.
- Only apply this step to the main image — restrict the step to just the feed's designated main image rather than the whole gallery.
- Acceptable quality — below this, fall back to the next step — if a step's result scores below the number you set, Renable moves on to the next step instead of keeping a bad result.
Drag a step to reorder it, and click one to change it. The order is the fallback order, so a step that only removes backgrounds cheaply belongs above the expensive one.
See Image background removal and classification for how the v1/v2/v3 models differ, why a typical pipeline chains a cheaper model first, what happens when nothing clears its quality bar, and how to review or override the result on a specific product.
The pipeline you build here is the catalog's default. A Remove image background rule can give some of the catalog's products a pipeline of their own — see Fixing your feed data with rules.
Locale#
Each catalog has a primary locale and fallback languages, which drive the language Aimée writes ad copy in and which customer reviews she quotes. See Languages and localization for details, including how number and currency formatting work.
Changes on this tab take effect when you Publish the catalog, and reach the catalog's products within minutes — you don't have to wait for the next import.
Prompts#
A table of AI prompts scoped to this catalog — these run on all of this catalog's products, regardless of which project is processing them. See Working with prompts for how to create and configure prompts.
Notes and events#
At the bottom of the catalog page, the Notes and events section logs feed import activity — successes, failures, parse errors, feed-item limit warnings — along with any notes you or your team add manually. Check here first if a catalog isn't behaving as expected.
Advanced settings#
An Advanced settings area holds a custom User-Agent header, for websites or CDNs that block automated scraping. See Questions about catalogs for when and how to use it.
Cost and limits#
AI enrichment features consume computing power, which is Renable's cost for running the service — some features, especially prompt-based copy generation, video generation, and the v2 background remover, cost noticeably more than others. Catalog count and product-per-catalog limits are also capped by your subscription plan. See Plans and limits and Understanding credits for details.