In short: Four weeks left. Content API for Shopping v2.1 shuts down on August 18, 2026 and there is no extension. This is the countdown plan for the time that remains. One grep to find every integration that will die, money paths migrated first, read paths second, and a parallel run so you switch the old API off on your own terms before Google does it for you.
The reminder fired this morning. I set it in June, the day I wrote about the Content API shutdown, and I placed it exactly four weeks before the deadline. On August 18, 2026, Google turns off Content API for Shopping v2.1 for good. Four weeks is enough time to migrate calmly. It is also enough time to convince yourself you will do it next week, four times in a row.
I already explained what breaks and why Merchant API is the better tool in the June piece. I won’t repeat it. This is the plan I run when a hard deadline gets close, one week per phase.
This week, find everything that will die
You cannot migrate what you haven’t found. Run this over every repo that touches Merchant Center:
grep -rEn \
"shoppingcontent\.googleapis\.com|ShoppingContent|content/v2(\.1)?([^[:alnum:]_.-]|$)|[\"']content[\"'][,[:space:]]+[\"']v2(\.1)?[\"']|\.content\([\"']v2(\.1)?[\"']" .The first pattern catches direct REST calls. The rest catch the official client libraries, because they hide the URL from you. Python builds the doomed client with build('content', 'v2.1'), PHP calls it Google_Service_ShoppingContent, Node uses google.content('v2.1'), and Apps Script projects load a ShoppingContent advanced service that never shows up in any repo. Check your script projects by hand.
Then put every hit in a sheet with four columns. What it does, how often it runs, who owns it, and how long until someone notices when it dies silently. That last column is the one you sort by.
Your grep only covers code you wrote. Zapier scenarios, feed tools, repricers and connectors can all sit on the old API too. Email every vendor this week and ask for their Merchant API date in writing. Their migration deadline is your outage.
Week two, migrate the money paths
Start where silence costs revenue. Price updates, stock updates, feed uploads. In Merchant API v1, GA since July 2025, these live in the products, inventories and datasources sub-APIs, and the compatibility guide maps the old calls to the new homes.
One thing surprised me pleasantly when we pointed our own tooling at the new endpoints at Lynt. The existing OAuth token with the old content scope just worked. Credentials carry over, code does not. You rewrite the calls and skip a new authorization dance with every client.
Week three, the read paths
Disapproval checks and reporting move next. This is also the week the migration starts paying you back. Reports page at 1,000 rows instead of 250, so the same job finishes in a quarter of the calls. And where your old code polled product statuses on a schedule, the notifications sub-API pushes changes to you instead.
Week four, switch it off yourself
Run both pipelines in parallel and diff the outputs for a few days. Then turn the old integration off about a week early, deliberately, while v2.1 still answers. If something you missed secretly depends on it, it surfaces while you can still flip the old calls back on for a day and fix it properly.
After August 18 there is no rollback. Google flips the switch for everyone at once, and whatever the grep didn’t find dies in production.
Twelve years on the Google Ads API taught me that hard deadlines are rare. Google usually extends, grandfathers, postpones. Not this one. So run the grep before you close this tab. The number it prints is the only migration estimate you need.
