A shopping feed fails in boring ways: a missing GTIN, a tax setting that doesn't match Merchant Center, a shipping attribute Google reads differently than you meant it. The go-live flow is built to catch those before Google does. Nothing ships until you press Go live.
The checklist
- 1Run guided setup. It walks the attribute mapping — shipping, tax, product identifiers (GTIN/MPN/brand) — and flags SKUs with missing or conflicting values before they become disapprovals.
- 2Open the preview. It renders exactly what Google will read, item by item: final titles, prices, availability, identifiers. What you see in the preview is byte-for-byte what ships.
- 3Fix what the preview surfaces. Missing identifiers and malformed shipping values are the two most common blockers — resolve them now, while the cost is an edit rather than a disapproval.
- 4Leave the feed in review mode as long as you need. Review mode is the default state; nothing is submitted to Merchant Center until the explicit Go live action.
- 5Press Go live. The first submission runs, and monitoring starts the same minute.
After go-live: Merchant Guardian
Disapprovals will still happen — policies shift and catalogs drift. Merchant Guardian watches for them continuously and auto-fixes the mechanical classes: policy flags, attribute problems, formatting issues. Each fix is a logged action, not a silent patch. What can't be auto-fixed is escalated with the specific disapproval reason attached.
Then: experiment
Once the feed is stable, feed A/B experiments test title strategies against each other — attribute order, keyword placement, length — resolved on click and conversion data rather than opinion. Titles are the highest-leverage text in the feed; test them like you'd test anything else that matters.
The preview exists so your first disapproval report is short. Read it like Google will, because Google will.