Catalog integration is the plumbing between your product data and every place a shopper might encounter your products: Google Shopping, Performance Max, Meta dynamic ads, Pinterest, and now AI assistants answering questions about what to buy.
Most furniture retailers have some version of it running. Far fewer have one that is actually working, and the failure is quiet. There is no error message when a feed silently drops half your catalogue. The ad spend continues either way.
The short answer
Your catalog is a structured file describing every product you sell: title, description, price, availability, image, brand, category, dimensions, GTIN where one exists. Catalog integration is how that file gets built, kept accurate, and delivered to Google Merchant Center, Meta Commerce Manager and anywhere else that needs it.
When it works, your products appear in Shopping results and dynamic ads automatically, and increasingly get quoted by AI assistants. When it does not, you are invisible in the highest intent placements on the internet while paying for text ads to compensate.
Why furniture breaks catalogs more than other categories
Most feed guidance is written for apparel and consumer electronics. Furniture violates several of its assumptions.
Variants explode. One sofa model in four fabrics, two orientations and three sizes is twenty four sellable items. Feeds that treat the model as one product lose every variant a shopper actually searches for. Feeds that publish all twenty four with identical titles get flagged as duplicates.
Most furniture has no GTIN. Barcodes are the assumed unique identifier across the feed ecosystem, and made to order furniture rarely has one. This is legitimate, and the fix is to declare `identifier_exists: false` correctly rather than inventing numbers, which is what a surprising number of feeds do.
Availability is not binary. Available now, floor sample, on order with an arrival date and special order are four different states, and a feed with only in stock and out of stock will misrepresent at least two of them. That is a trust problem with the platform and with the buyer, and it is covered in more depth in our piece on inventory management for furniture retailers.
Dimensions matter more than almost anything. Will it fit is the single most common pre purchase question in furniture, and width, depth and height belong in structured product data rather than buried in a description paragraph.
Shipping is genuinely complicated. Freight, white glove delivery, assembly, delivery radius and surcharges by zone. A feed with a flat shipping value will misprice a large share of your catalogue.
The four things a furniture feed usually gets wrong
Titles written for humans instead of for matching. "The Aspen Collection" tells a shopper nothing and a matching algorithm less. A title that works reads brand, then product type, then defining attributes: "Ashley Aspen 3 Seat Reclining Sofa, Grey Fabric, 88 inch". Unglamorous, and it wins.
Descriptions copied from the manufacturer. Every retailer carrying that line submits the identical text. Rewriting descriptions is one of the few places where effort separates you from other stores selling the same item, and it is the same work that improves ecommerce SEO for furniture.
Images that are not what the platform wants. White background primary image, no watermarks, no promotional text burned in, high resolution. Lifestyle shots are excellent on your site and on social, and they get product listings disapproved when submitted as the primary image.
A sync frequency slower than the business. Weekly feeds on a catalogue that changes daily are wrong most of the time by design. Daily is a floor. Live is better.
How the pieces actually connect
Your product data lives in one system of record, usually the website platform or an inventory system. From there it is transformed into a feed and delivered to each destination, and each destination wants a slightly different shape.
Google Merchant Center feeds Shopping, Performance Max and free product listings. Meta Commerce Manager feeds dynamic ads and shop surfaces on Facebook and Instagram. Both accept scheduled fetches, direct uploads or an app integration.
The important architectural point is to have one source of truth and transform outward, rather than maintaining separate catalogues per platform. Retailers who edit the Google feed and the Meta feed independently end up with two products that disagree about price, which is exactly the inconsistency that causes disapprovals.
If your site is on Shopify, most of this is configuration rather than engineering. If it is not, the integration is usually the honest reason to consider a rebuild, which is part of what website and conversion work covers.
Catalog integration is now an AI visibility question
This is the part that has changed recently and that most retailers have not adjusted to.
AI assistants answering "best sectional under 2000 dollars" or "who sells recliners near me" draw on structured product data, not only on written pages. Structured, accurate, well described product data is what makes your inventory quotable. Vague titles, manufacturer boilerplate and wrong availability make it unusable, and an assistant will simply name a competitor whose data is clean.
The overlap is convenient: the same work that fixes Shopping performance also determines whether an AI assistant can recommend you. Feed hygiene has quietly become an AI search project.
How to check whether yours is actually working
Open Merchant Center and compare the number of active products against the number of products you sell. A gap is the whole story, and a gap of more than a few percent means something is being rejected silently.
Then read the disapproval reasons rather than the count. Missing GTIN, image quality, price mismatch and availability mismatch account for most furniture rejections, and each has a specific fix.
Then check the price on the feed against the price on the site for ten products picked at random. Mismatches are the most common cause of a whole account getting restricted, and they usually come from a sale price the feed did not pick up.
Finally, look at how many of your products have ever received an impression. A catalogue where a small handful of items absorb all the delivery usually has a titling problem across everything else, not a bidding problem.
If you would rather have this run properly alongside the campaigns, that is part of paid ads management and Google Ads for furniture stores.
Frequently asked questions
What is catalog integration?
It is the process of connecting your product data to the platforms that display it, mainly Google Merchant Center and Meta Commerce Manager, so your products can appear in Shopping results, dynamic ads and AI generated answers without being listed by hand.
Do I need a product feed if I only run search ads?
Not strictly, but you are leaving the highest intent placements unused. Shopping and Performance Max both require a feed, and they typically reach buyers closer to purchase than text ads do.
My furniture has no barcodes. Can I still run Shopping ads?
Yes. Set `identifier_exists` to false for those products, which is the supported way to declare it. Do not invent GTINs, which risks the whole account.
How often should the feed update?
Daily at minimum for price and availability. If stock moves faster than the feed, you are paying for clicks on listings that are already wrong.
Should each fabric and size be a separate feed item?
Yes, as variants of a parent product, with distinct titles and images. That is how a shopper searching for a specific configuration finds it rather than a generic model page.
Why are my products disapproved for image quality?
Usually a lifestyle photo submitted as the primary image, or promotional text and watermarks burned into it. The primary image needs a plain background and no overlays. Lifestyle shots still belong on your own product pages.