Negative Keywords
Queries you tell Google Ads to never serve your ads against. The structural defence built from this report.
Read the entry →Stan Consulting · Marketing Atlas · Reference · Google Ads
The Google Ads report that names the actual queries your ads served. The most-ignored diagnostic surface in any account.
Section 02 · Quick definition
The search terms report is the Google Ads report that lists the actual user queries that triggered your ads to serve, alongside the keyword and match type that captured each query. It is the bridge between what an operator asked Google to bid on and what Google actually bid on. The report updates daily, supports filtering by campaign and ad group, and is the primary surface used to identify match-type drift, branded leakage, missing negatives, and queries with no commercial intent. Without reading it, an account is bidding on terms its operator has never seen.
Section 03 · Why it matters
The search terms report is the only place in Google Ads where the gap between intended targeting and actual targeting becomes visible at the query level. Every other surface aggregates: campaign rollups, ad-group rollups, keyword views. The report breaks the aggregate apart and names the queries individually. Without it, an account's actual bidding behavior is invisible.
It also produces the second-order outputs the rest of an audit depends on. Match-type drift cannot be measured without it. Branded leakage cannot be quantified without it. Negative keywords cannot be added intelligently without it. Conversion-signal corruption is harder to detect without it. The report is upstream of most structural fixes.
For an operator, the practical stake is that most of the budget is allocated against queries the operator has never read. Reading the report flips the relationship: the account starts bidding against queries the operator has approved, not against queries the algorithm and the match-type rules surfaced on their behalf.
Section 04 · How it works
The search terms report aggregates query data Google Ads collects on every served impression. Privacy rules constrain what gets shown, but the bulk of the spend appears at the query level for any account with sufficient volume.
Each row represents one served query, with the keyword that captured it, the match type the keyword used at the time, plus impressions, clicks, cost, conversions, and conversion value over the selected window.
Queries with low search volume and queries Google considers personally identifying are aggregated under the “Other search terms” row. The percentage of spend that lands there is itself a diagnostic signal.
Operators can add a query as a keyword (positive) or add it as a negative directly from the report row. Both decisions update the account immediately and propagate through the bidding model on the next learning cycle.
The “match type” column shows which match type captured each query. Reading the column reveals which keywords are over-expanding into adjacent intent, the core mechanism of structural waste.
The data updates daily but is interpreted at trailing 30- to 90-day windows for stability. Weekly reading is operationally heavy; monthly is the realistic cadence for most accounts.
Section 05 · Common misunderstandings
“The report shows every query that triggered an ad.”
It shows the queries above the privacy and volume thresholds Google sets. The remaining spend lands in the aggregated “Other” row. The fraction of spend in that row is itself a number worth tracking.
“It only matters for broad-match campaigns.”
Phrase match and even exact match expand into close variants. The report is the only place that variant expansion becomes visible. Ignoring it for non-broad campaigns leaves the same gap, smaller.
“Smart Bidding makes the report obsolete.”
Smart Bidding decides what to bid; the report decides what to bid against. The two surfaces answer different questions. Smart Bidding does not negative-list a bad query on the operator's behalf.
“Reading once a quarter is enough.”
Quarterly cadence accumulates a quarter of waste before review. Monthly is the operating floor for live accounts; weekly is the floor during a restructure or a launch.
Section 06 · Diagnostic questions
When was the report last opened, and by whom?
What share of last 90 days' spend appears in named queries versus the aggregated “Other” row?
Of the named queries, what percentage are queries the operator can affirm as commercial intent?
How many distinct queries are pulling spend from each high-traffic keyword, and which match types are involved?
Which queries are converting against ad groups they semantically do not belong to?
What is the cumulative spend on branded queries served by non-branded campaigns over the last 90 days?
What review cadence and ownership exist for the report inside the operator's team?
Section 07 · Related Atlas entries
Section 08 · Five Cents
The search terms report is the report nobody opens until somebody else opens it for them. The cultural reason is mundane: it is long, it is unranked, and it does not have a chart. A pivot table that names the queries you served is the most useful object in the account, and it shows up looking like a parking ticket. So it sits closed. Then the audit gets ordered, the report gets opened, and the operator sees a third of the spend going against queries they would have screamed at if anyone had read them out loud. That is not an exaggeration. That is the median first-look.
Stan · Marketing AtlasSection 09 · Sources
Continue in the Atlas