Content API for Shopping is not being deprecated, scheduled for removal or entering a transition period. It is already gone. Google’s documentation states it plainly: "Content API for Shopping was sunset on August 18, 2026."
That was Tuesday. If something in your product feed pipeline started failing this week and nobody connected it to a calendar entry, this is the first place to look.
What actually happened, and when
The Merchant API documentation carries the sunset notice in the past tense, which is the tell that this is a completed event rather than an upcoming one.
Merchant API is the successor. Google’s framing is a migration rather than a replacement in kind, inviting developers to "Compare with the Content API for Shopping and learn how you can migrate to the Merchant API."
The distinction matters operationally. A deprecation gives you a warning and continued service. A sunset stops the service. Anything still calling the old endpoints is not degraded, it is finished, and the failure will show up as feed staleness rather than as a loud error in a place anyone monitors.
The line most coverage will miss
Here is the part worth knowing today, and it is sitting in Google’s own docs.
The sunset is not absolute. The same documentation says: "If you need additional time to migrate to Merchant API, apply for extended access to Content API for Shopping."
So the accurate statement is not that Merchant API is now the only programmatic route into Merchant Center. It is that Merchant API is the only route unless you apply and are granted extended access. That is a meaningful difference for anyone who missed the date, and it is the kind of detail that gets flattened out of secondary coverage into a cleaner and less useful sentence.
Google publishes no public conditions or deadline for that extended access in the documentation available at the time of writing, so treat it as a request rather than an entitlement, and apply early rather than assuming it will be there indefinitely.
Who is actually exposed
Most merchants will never notice this, and it is worth being clear about who will.
You are probably fine if your products reach Merchant Center through Shopify, WooCommerce, BigCommerce, a major feed management tool, or an agency that handles the integration. Those vendors migrate on their own schedule and generally did so months ago.
You are exposed if someone at your company wrote a direct integration, if a legacy internal system pushes inventory or pricing on a schedule, if a small development shop built your feed years ago and has since moved on, or if a plugin you installed and forgot about is doing the talking.
That last category is the dangerous one, because nobody owns it. The integration works, so it is invisible, and it stays invisible right up until prices stop updating and nobody can say when they last did.
The failure mode is silence, not an error
This is the practical reason a passed deadline is worse than an upcoming one.
A broken feed integration rarely announces itself. Your listings do not vanish. Merchant Center keeps serving whatever it last received, so products stay live with stale prices, stale stock levels and stale availability. You keep paying to advertise items you may no longer have at prices you may no longer honor.
The visible symptoms arrive later and look like something else entirely: rising disapprovals for price mismatch, a climb in out-of-stock complaints, conversion rates drifting down without an obvious cause. None of that reads as an API problem in an analytics dashboard. Anyone reading those signals will reach for bidding, creative or landing pages first, which is the same diagnostic trap we described when Google expanded Limited Ad Serving to all campaigns. The constraint is upstream of everything you would normally adjust.
Migrating off Content API for Shopping is a scoping question
Before anyone estimates this, work out what you were actually using, because the answer varies by an order of magnitude.
Google frames the move as a migration with a comparison guide rather than a swap of endpoint URLs, which is a reasonable signal that the surface differs enough to need real work rather than a find-and-replace. What that work costs you depends entirely on the surface area you touched.
A pipeline that only pushed a product catalog on a schedule is a contained job with a clear test: the same products appear with the same attributes. A pipeline that also handled inventory synchronization, promotions, regional pricing, shipping settings or reporting is several jobs, each with its own edge cases, and each capable of failing quietly in the way described above.
The sequencing advice is the same one that applies to any forced migration. Establish what the current integration actually does, in writing, before you touch it. Teams routinely discover that a feed script has been doing something nobody documented, and the discovery is much cheaper before the rewrite than during it. The same reasoning applies to how you measure the outcome: if you cannot state what the old pipeline produced, you cannot tell whether the new one is producing it.
Two Google tightenings in eight days
Worth putting these next to each other, because they are not unrelated.
On August 5, Google published the update expanding its Limited Ad Serving policy to cover all Google Ads, running gradually through 2028. On August 18, Content API for Shopping was sunset in favor of Merchant API.
One governs whether your ads serve. The other governs how your products get in. Both replace a long-standing arrangement with a newer, more tightly specified one, and both are the sort of change that produces silent degradation rather than an outage. If you run commerce on Google, the surface area you need to actively monitor got larger this month.
What to check this week
Five positions, in the order I would take them.
Find out who owns your feed integration. Not which tool, which person. If the honest answer is that nobody currently owns it, that is your actual finding and it matters more than the API change.
Compare Merchant Center data against your source of truth. Pull ten products and check price, availability and stock against your own system. A stale feed looks completely normal in the interface, so this comparison is the only reliable test.
Apply for extended access if you are broken. Google documents the path. Use it to migrate deliberately rather than rewriting an integration under pressure.
Ask your platform or agency for the migration date in writing. Most did this quietly. Getting the confirmation costs one email and closes the question permanently.
Diary the surrounding deadlines. Limited Ad Serving rolls out through 2028, and this is unlikely to be the last such change. Reviewing programmatic dependencies on a schedule beats discovering them through a symptom, which is the same argument we made about measurement infrastructure becoming something you maintain rather than configure once.
The Merchant Center Help hub is the reference for the account side. The developer documentation is the reference for the API side. Both are worth a look this week rather than next.
Frequently Asked Questions
When was Content API for Shopping shut down?
Google’s Merchant API documentation states that Content API for Shopping was sunset on August 18, 2026. The wording is past tense, which means this is a completed change rather than an announced future one. Any integration still calling the old endpoints is no longer supported as of that date.
Is Merchant API the only way in now?
Almost, but not quite, and the difference matters if you missed the date. Merchant API is the successor and the documented path forward. However, Google’s own documentation also says that if you need additional time to migrate, you can apply for extended access to Content API for Shopping. So the accurate statement is that Merchant API is the only route unless you apply and are granted an extension.
How do I know whether I am affected?
Establish how your products reach Merchant Center. If it is through Shopify, WooCommerce, BigCommerce, a feed management platform or an agency, that vendor almost certainly migrated already. If a custom integration, an internal system, or an old plugin talks to Google directly, that is where the risk is. The practical test is comparing prices and availability in Merchant Center against your own system for a sample of products.
What does a broken feed look like?
Usually like nothing at all, initially. Listings stay live and Merchant Center keeps serving the last data it received, so products remain visible with stale prices, stock levels and availability. Symptoms surface later as price mismatch disapprovals, out-of-stock complaints and drifting conversion rates, all of which look like marketing problems rather than an integration problem.
Can I still get an extension?
Google documents an application for extended access for those who need more time to migrate. No public conditions or closing date appear in the documentation available at the time of writing, so treat it as a request that may be granted rather than a guaranteed entitlement, and apply sooner rather than later if you need it.
How big a job is migrating?
It depends entirely on how much of the old API you used. A simple product upload integration is a contained piece of work. Something that also handles inventory updates, promotions, regional pricing or reporting is a larger project. Google positions the change as a migration with a comparison guide rather than a drop-in replacement, which is a reasonable signal that the surface differs enough to require real work.
Should small merchants worry?
Only if they have a custom integration, which most do not. The risk correlates with how bespoke your setup is rather than with company size. A small merchant on a mainstream platform is safer than a mid-sized one whose developer built something clever in 2019 and has since left. The question to ask is not how large you are but how unusual your pipeline is.
Where should I look first if something is wrong?
Merchant Center diagnostics for the account, then your own logs for authentication or endpoint failures around August 18. If those confirm the integration stopped, apply for extended access before starting emergency work, then plan the migration properly. Rebuilding a feed pipeline in a hurry tends to produce a second problem shortly afterwards.
