Run the negative audit backwards. Pull every negative list attached to the account and check its terms against the account's own converters across multiple windows — greedy single-token negatives on shared lists are the ones doing damage.

On this page

• Why audit negative keywords in the first week of a takeover?
• How do I audit inherited negative keywords?
• Which negative keywords do the most damage?
• Can a zero-conversion term convert in a longer window?
• What did I get wrong the first time?
• What evidence bar should a negative keyword decision meet?

Why audit negative keywords in the first week of a takeover?

Negatives are invisible. Nothing in the interface tells you a phrase stopped converting because someone added a broad token two years ago. The traffic simply is not there, so it never appears in any report you would think to run.

How do I audit inherited negative keywords?

1. Enumerate every negative source. Campaign-level lists, ad-group-level negatives, and — most importantly — shared lists, because one greedy token on a shared list applies everywhere at once.
2. Pull converting search terms across several windows spanning about a year. Last 30 days, last

90, and 3 prior 90-day blocks.

1. Match negatives against converters. Any negative that matches a term which converted in any window is a suspect.
2. Check phrase and broad negatives against supersets. A phrase negative on a category word kills every longer query containing it, including the converting ones.

Which negative keywords do the most damage?

Single-token broad negatives. how, best, where, order, free, cheap, reviews.

On a telehealth account, June 2026, an audit's "safe" negative list included how to get, for sale and reddit. All three had converted — 6.5, 1.3 and 1.1 conversions across 90 days. The audit had used a 14-day window, which hid every one of them.

Can a zero-conversion term convert in a longer window?

A search term read $224 spent, 0 conversions at 30 days. On the standard advice it gets negated.

Pulled at 90 days: $894 spent, 9 conversions, $99 cost per acquisition. One of the better terms in the account.

A single window hides converters. Even 90 days hides converters. A 5-window full-year re-test on that same account dropped 3 more terms from a list that had already survived a 90-day check.

What did I get wrong the first time?

On one negatives pass I ruled five terms as new Tier 3 candidates — "do not negate" — without ever checking the shared lists that were already holding them. The recommendation was coherent and the work was wasted, because the decision had been made years earlier by somebody else.

Reconcile against the shared lists before analysing anything. Find out what is already blocked before you reason about what should be.

What evidence bar should a negative keyword decision meet?

• Sizing the waste — one window is fine. You are estimating.
• The kill/keep decision — multi-window is mandatory. This is the gate.

Related questions

I'm getting clicks from Google Ads but no conversions. Is it the traffic or my landing page?

You cannot tell from the account-level number, and almost every published answer tells you to fix both at once. Segment conversion rate by search-term intent instead. Read the full answer.

How long should I wait before deciding whether a Google Ads change actually worked?

Two clocks run at once and almost nobody separates them. One is whether the bid strategy has adjusted. Read the full answer.

My Google Ads campaign isn't spending its daily budget — why, and what do I check first?

Check budget utilisation against impression share lost to rank before anything else. An underspending campaign is either strangled by its target, rank-locked by its bid ceiling, or genuinely out of demand — and the three have opposite fixes. Read the full answer.

Written by Usama Khan, who runs Google Ads for e-commerce brands and home-services companies at AdventurAds. Figures are from accounts under active management and are anonymised. Last reviewed 2026-08-26.