What free listings are, and what they are not
Google Merchant Center started as the pipe that fed Shopping ads. Since free listings arrived it is also the pipe that feeds a set of unpaid product placements, and most Indian D2C brands we look at are either not using it or using it badly enough that half the catalogue is disapproved without anyone noticing.
Google's own documentation lists the surfaces: Search, including rich results and the Popular Products carousel, the Shopping tab, Images, Lens, YouTube, Maps, Business Profile and Gemini. In most accounts the feature is switched on by default once products are flowing.
It's worth being precise about what this is, because it gets oversold in both directions.
- It is not advertising. No bid, no budget, no media cost. The cost is entirely the hours to build and maintain the feed.
- It is not a ranking shortcut for your website. Free listings and organic blue links are different pipelines. A product can appear in the Shopping tab while its product page ranks nowhere, and the reverse is just as common.
- It is not guaranteed placement. Eligibility is what you earn. Which products actually show, and where, is Google's call and varies by query.
- It is not a substitute for ecommerce SEO. Category pages, buying guides and internal links still do the work of getting found by people who don't yet know what they want. The feed serves people who already do.
- It is not only for large catalogues. A twelve-SKU brand can run this. Fewer products actually makes the feed easier to keep clean.
The feed decides whether a product appears at all
The feed is a structured description of each product. Google's product data specification is the authority on what each attribute means, and the required set is short enough that there's no excuse for getting it wrong.
| Attribute | Requirement | Where Indian sellers go wrong |
|---|---|---|
id | Required, unique, max 50 characters | Regenerating IDs on every export, which resets each product's history. |
title | Required, max 150 characters | Marketing names with no attributes. Front-load brand, product type, then colour and size — that's what people type. |
description | Required, max 5,000 characters | Copy-pasted from the manufacturer for every variant, so nothing distinguishes them. |
link and image_link | Required | Images under 500x500, or a hero shot with a festive-offer badge burned into it. |
price | Required, ISO 4217 currency | Feed price excludes GST and the landing page includes it. Google requires the price to match the landing page, the structured data and checkout. |
availability | Required | Nightly feed while stock moves hourly. The listing says in stock, the page says sold out. |
brand | Required for almost all new products | Left blank by unbranded sellers. Use your own brand if you are the manufacturer. |
gtin / mpn | GTIN strongly recommended where it exists; MPN required if there is no manufacturer GTIN | Inventing or reusing a GTIN. Google is explicit — only supply one if you're sure it's correct, because a wrong one gets the product disapproved. |
maximum_retail_price | Optional, available for India | Submitting an MRP that doesn't match the one printed on the page. If you show MRP, it has to agree. |
The disapprovals that hit Indian sellers hardest
Item-level issues take a product out. Account-level issues take everything out. In roughly the order we find them:
- Price mismatch. Almost always GST. Decide once whether your catalogue is inclusive or exclusive, make the feed, the page, the structured data and the checkout agree, and stop the argument permanently.
- Availability mismatch. Fix the sync frequency, not the individual products. If your feed refreshes once a night and you sell out at 11am, you will keep failing this.
- Image rejections. Google requires at least 500x500 and recommends 1500x1500 or above, and it prohibits promotional elements, calls to action, price stickers, watermarks and borders. Indian D2C creative is festival-driven and full of badges, so this bites hard around sale season. Keep one clean product shot for the feed and put the badges on the site.
- Bad or invented GTINs. A borrowed barcode from a similar product will get caught. If you genuinely don't have one, say so with the identifier fields rather than guessing.
- Missing contact route or return policy. Google's checkout requirements ask for a way to contact you — a form, an email, a phone number or a social profile — and for refund and return information that's easy to find during checkout. A contact page that's only a WhatsApp button embedded in an image doesn't read as a contact route.
- Insecure checkout. A valid SSL certificate across the sign-up and checkout flow, not just the home page. Mixed-content warnings on a payment step will hold up the whole account.
Feed data and on-page structured data have to agree
There are two ways to tell Google about a product: the feed, and Product structured data on the page itself. They are not alternatives. Google's documentation says that providing both maximises your eligibility to product experiences, and that some experiences combine the two — pricing can come from the feed when the page's markup doesn't carry it.
Which means the failure mode is disagreement, and disagreement is normal on Indian storefronts because the two systems are usually owned by different people. The feed comes out of an inventory export the ops team maintains. The structured data comes out of a theme or a plugin the developer installed in 2023. Nobody has compared them since.
The checks worth running once a quarter:
- Price — feed, on-page markup, visible page price and checkout price all identical, including whether GST is inside or outside the number.
- Availability — the markup says
InStockwhen the page says in stock, and both change when the warehouse does. - Currency —
INReverywhere, including on any international variant of the page. - Identifiers — the same GTIN or MPN in the feed and the markup, or absent from both.
- Canonical URL — the feed's
linkpoints at the canonical product URL, not a filtered or session-parameter version. If you're unsure what that means, canonical tags covers it. - Variants — one entry per purchasable variant, each with its own image and price, rather than one parent row that hides six colours.
The hours it takes, and why we fold it into a retainer
We publish our prices, so we can be blunt about this rather than quoting it as a separate line with an impressive-sounding name.
A first setup — account, website claim, shipping and returns configuration, feed source, an initial pass on disapprovals — is roughly 8 to 12 hours for a catalogue in the tens to low hundreds of SKUs. A very large or badly structured catalogue takes longer, mostly because the export needs work upstream rather than because the feed does. After that, maintenance runs about 1 to 2 hours a month: check the issues list, fix whatever the last theme update broke, review titles for the products that matter.
That fits inside the six technical hours a month in our ₹75,000/mo SEO plan, so we do it as part of the work rather than selling it. We're not being generous — it's the honest place for it. Charging ₹25,000 for a Merchant Center setup would be charging a day's work at agency day-rate for something that is a genuinely small task, and the ongoing hour or two a month is cheaper to absorb than to invoice. Our pricing is on the pricing page, and the smaller-site plan at ₹40,000/mo covers it too, at the cost of some content cadence.
The reason to do it at all is the ratio. Ten hours of setup against a surface your competitors are largely ignoring is unusual arithmetic in a channel where most gains cost a quarter. It will not carry a business on its own — free listings serve people who already know what they want, and the volume follows demand for your category rather than your effort. But it is the cheapest incremental visibility available to an Indian D2C brand right now, and most brands still haven't claimed it.