Google's Content API Sunset Is Breaking Feeds Now

Google's Content API for Shopping started throwing errors on September 1. Not a warning. Errors, live, on calls that were working fine three weeks ago. If your product data reaches Google Shopping through anything other than Shopify's own a

Google's Content API for Shopping started throwing errors on September 1. Not a warning. Errors, live, on calls that were working fine three weeks ago. If your product data reaches Google Shopping through anything other than Shopify's own app, there's a real chance your feed is degrading right now and nobody on your team has noticed yet.

Here's what actually happened, who it hits, and the five-minute check that tells you which side of it you're on.

Google set a deadline in 2025 and just started enforcing it

Google announced it would retire the Content API for Shopping back on August 19, 2025, giving developers a year's notice before the August 18, 2026 shutdown. That's the kind of deadline everyone means to get to eventually. Google's own release notes for the Content API confirm what happens to the stragglers: starting September 1, 2026, requests against the old API "will experience progressive errors." Not a hard cutoff on day one. A slow bleed that gets worse the longer it runs.

The replacement, the Merchant API, has been generally available since summer 2025. It isn't a drop-in swap. Resource identifiers changed, price fields changed from a value:string to an amountMicros:int64 format, and the old customBatch method Google removed entirely. Anyone who built a direct integration against the old endpoints has real code to rewrite, not a settings toggle to flip.

This deadline lands harder than a typical API version bump because of what runs on top of it. Performance Max, the format most Shopping budgets flow through now, doesn't let you pick individual products to promote the way old-style Shopping campaigns did. It reads whatever Merchant Center says is live and eligible, and builds the campaign from that. A feed quietly losing sync doesn't shrink the campaign gracefully. It starves a black box that was already hard to audit, and the first sign most owners get is a cost-per-conversion that's drifted for reasons nobody can point to.

Most Shopify stores are fine. Some aren't, and won't know it

Here's the part the panic headlines skip. If you sync products through Shopify's built-in Google & YouTube channel app, Google's own guidance is explicit: merchants using a technology partner like that app "don't need to do anything," because the provider handles the API migration on the back end. Merchants discussing the rollout in Shopify's own community forum report the same thing in practice, including product IDs and conversion history carrying over untouched.

That covers a lot of stores. It doesn't cover all of them.

The businesses most likely to be exposed are also the ones least likely to notice quickly: a growing store that outgrew the default app a while back, an agency-built integration nobody's touched since launch, or a multi-storefront brand syncing feeds through custom middleware. None of that shows up as a dramatic outage. It shows up as a slow decline in impressions that looks like a market problem until someone checks the actual API calls.

Sellers running on multiple marketplaces at once are a particular blind spot here. A brand pushing the same catalog into Google, a marketplace like Etsy or Walmart, and its own site through a third-party feed management tool is trusting that tool's migration timeline, not Google's. If that vendor is still finishing its own move to the Merchant API, every store it serves inherits the delay, and most of those stores have no direct visibility into whether their specific vendor has actually finished.

The stores that get hit hardest by deadlines like this are never the ones running the simple setup. They're the ones with a custom integration built two years ago by someone who's since moved on, quietly working right up until the day it doesn't.

— the pattern we keep finding in client feed audits

The five-minute check, and the actual cost of skipping it

If you or your developer manage the feed integration directly, the fastest diagnostic is the request URL itself. Calls to shoppingcontent.googleapis.com mean you're still on the old Content API. Calls to merchantapi.googleapis.com mean you've already migrated. If nobody on your team can answer that question in under a minute, that's the real signal: you don't have visibility into a system your Shopping ads depend on.

The stakes are not abstract. A disapproved product in Merchant Center earns zero impressions and zero revenue from Google's Shopping surfaces, according to Google's own product status documentation. Nothing about that failure is loud. Your Performance Max campaign doesn't pause itself and send an alert. It just quietly serves fewer products, spends the same budget, and produces worse numbers that look like a demand problem until someone checks the feed.

For a Shopify store that only ever installed the official Google & YouTube app, this whole post is good news you can skip past: you're already covered, per Google's own Merchant Center guidance on third-party technology partners. For anyone running something more custom, and a meaningful share of growing ecommerce businesses are, this is worth a direct check this week, not a someday-task on a backlog.

There's a second diagnostic worth running alongside the URL check: open Merchant Center itself and look at the Diagnostics tab for your product data source. A source still reporting through the old system tends to show stale "last updated" timestamps that don't move even after you've pushed catalog changes, or a sudden cluster of new item-level issues that weren't there a month ago. Neither symptom announces itself as "your API is sunsetting." Both look, at first glance, like a data-quality problem you'd normally chase in the wrong direction.

Migrating is a rewrite, not a checkbox

If your integration is on the old API, the fix isn't cosmetic. The resource model changed, the pricing fields changed format, and batch operations work differently. Google's migration compatibility guide lays out the field-by-field differences, and treating this as a find-and-replace job on the old endpoint usually produces a feed that half-works, which is worse than one that obviously doesn't.

Google does offer breathing room: an extended-access request form for teams still mid-migration, approved on submission if you supply the correct Google Cloud project IDs. It's a genuinely low-friction ask, worth filing the same day you realize you're behind. But it's a stay of execution, not a fix. It buys the weeks your developer needs to do the actual rewrite. It doesn't rewrite the integration, restore the feed rules and supplemental feeds tied to the old data source, or catch the smaller platform-specific gotchas, like a supplemental feed override that silently stops applying, that tend to surface mid-migration rather than on day one.

This is the same shape of deadline we flagged in Shopify's checkout extensibility cutover: a platform quietly retires the old way of doing something, most merchants are covered by default, and the ones who built something custom find out the hard way. If your store runs on a heavily customized Shopify Horizon build or leans on Performance Max and other automated Google Ads formats, you're exactly the profile with the most custom surface area for a migration like this to quietly break.

Nobody budgets time for a deadline they didn't know existed

This is the third platform deadline we've tracked this quarter that works the same way: announced quietly, enforced on schedule, and invisible to most business owners until something downstream stops converting. WhatsApp's own pricing shift and Shopify's checkout cutover both followed the identical pattern. The API changes on a date nobody put on their calendar, the businesses running the default setup are fine, and the businesses running something custom find out three weeks later when a number that used to be predictable stops being predictable.

The honest version of this advice: check the request URL yourself if you can, today. It takes five minutes and costs nothing. If you can't answer it, or if the answer comes back "we're not sure who built that," that uncertainty is the actual risk, more than the API sunset itself. We audit and migrate product feed integrations as part of the same work we do keeping a store's ad tracking and automation stack current, because a feed that quietly stops updating is functionally the same problem as a pixel that quietly stops firing: money spent against numbers nobody's actually watching.