Skip to main content
FHAVI

Free resource

Fix one item properly before you touch nine hundred

Bulk-editing a catalogue to clear disapprovals is how a small data problem becomes a large one. The triage sheet exists to make one item pass first — because the fix you learn from it is the fix you apply in bulk, and the fix you guess at is the one you apply twice.

The order of work

One item, then the pattern, then the batch

01

Do not skip step 2

  1. Pick one disapproved item. Open its detail panel, not the summary list — the detail names the attribute or policy that failed.
  2. Classify it: data error (a field says something untrue), policy issue (the product or claim is restricted) or destination problem (the landing page disagrees with the feed). The three have different fixes and no overlap.
  3. Fix only that item's cause, then request a re-review of that item alone.
  4. Wait for the result before touching anything else. A pass confirms your diagnosis; a second disapproval means the cause was upstream.
  5. Group the remaining items by the same reason, and fix them as a batch in the source data rather than row by row in the feed.
  6. Re-check that the disposal rate falls before requesting an account-level review — if items are still failing, that request is spent on a known-broken state.

Classification

Where each class actually lives

02

Data error

  1. Price, availability, identifier or image — fix in the feed source, then let the next fetch carry it.
  2. Check whether the source of truth is the store or a supplier sheet; fixing the feed while the store disagrees produces a repeat next week.
  3. Verify the fix on the landing page, not only in the feed preview.
03

Policy issue

  1. Read the specific policy named, and check whether it applies to the item or the whole category.
  2. Trademark: remove the borrowed brand from title, description and image — not just the title.
  3. Restricted categories: either supply the approval and disclosures, or remove the items. Editing copy does not clear a policy.
  4. Repeated identical violations escalate from item-level to account-level action, so a repeated reason is urgent.
04

Destination problem

  1. Open the feed link in a private window: does it load the exact product at that price?
  2. Check for redirects that strip parameters, login walls, or a page that renders empty to a crawler.
  3. Confirm checkout is available in the target country with a working payment method.
  4. These look like feed failures in the summary and are never fixed by editing the feed.

Recording

The columns that stop you repeating this

05

Per item

  1. Item id and title, so the row is identifiable after the fix.
  2. Reason exactly as Google wrote it — copy the string, do not paraphrase.
  3. Class: data, policy or destination.
  4. The field or page that actually failed.
  5. What was changed, in one line.
  6. Re-review result and date.
06

Per week

  1. Count of disapproved items, and the change from last week.
  2. The top three reasons, by count, not just the total.
  3. How many fixes came from the source data versus patched in the feed — the ratio tells you whether the process is improving.

More resources

Keep going

The rest of the library

Every checklist here is free and ungated. The research pages carry the reasoning behind them.

All resources · Research library

Disapprovals coming back after every fix?

That usually means the source data is wrong, not the feed. Tell us where the data comes from and we will say what to change upstream.

Prefer email? Write to md@fhavi.com