Search terms that never convert: an audit walkthrough

The keywords you choose are your best guess at what buyers will type. The search terms report shows what they actually typed. The gap between the two is where budget quietly disappears, into queries you never meant to bid on and would never choose by name. This walkthrough turns that report into a shorter bill.

Keywords vs search terms: what is the difference?

A keyword is the phrase you bid on. A search term is the real query a user typed to trigger your ad. With broad and phrase match, one keyword can match hundreds of search terms, some spot-on and some far off. The report lists the real terms, with the spend and conversions each produced. That is the ground truth of where your money went.

The audit, step by step

1. Pick a fair date range. Choose a window long enough to gather real data but recent enough to reflect how the account runs now. Ninety days suits most accounts; a high-volume account can use less.
2. Sort by spend, filter to zero conversions. Order the report by cost, then filter for terms with no conversions. The top of that list is your suspect pile: the terms that took the most money and gave nothing back.
3. Read intent, not just the words. A term can share your keywords and still mean something else. Someone searching for a free version, a job at your company, or a rival by name is not your buyer. Judge what the person wanted, then decide.
4. Separate irrelevant from unprofitable. Two different problems hide here. An irrelevant term should become a negative keyword. A relevant term that simply costs too much is a bidding or landing-page problem, not a query to block. Do not negative your way out of a conversion problem.
5. Add negatives at the right level. Place clear negatives at the ad group, campaign, or account level depending on how broadly the term should be blocked. Build shared lists for terms that never apply anywhere, like job seekers or free-tool hunters.

Suppose one term spent $300 across the quarter with no conversions and reads as clearly off-intent. That is a clean negative, and $300 a quarter back in your pocket.

How do you keep the list clean?

A one-time cull helps for a month, then broad match finds new terms. Make the review a habit. A short weekly pass over new high-spend, zero-conversion terms keeps the waste from rebuilding. Save your negative lists and reuse them across campaigns so the same bad query never costs you twice. This audit pairs with the wider wasted-spend audit and with finding budget-capped winners to fund.

Which patterns show up again and again?

After a few audits, the same kinds of wasteful term keep appearing. Learning to spot them by pattern turns a slow line-by-line read into a fast scan.

• Job seekers. Searches with careers, jobs, salary, or hiring want employment, not your product. They rarely convert and belong on a permanent negative list.
• Free and DIY. Queries with free, template, or how-to-do-it-yourself signal someone who does not intend to pay. Relevant to your topic, wrong for your ad.
• Competitor names. A user searching a rival by name usually wants that rival. Bidding there can work, but treat it as a deliberate strategy, not an accidental broad-match spill.
• Wrong meaning. A keyword with two senses pulls in the sense you do not sell. A term for the other meaning is a clean negative.
• Too broad to judge. One-word queries that could mean anything spend without telling you intent. Tighten the keyword or add qualifiers rather than blocking outright.

Keep a running list of these as shared negative lists, and each new campaign starts cleaner than the last. The point of the audit is not to win once. It is to stop paying for the same mistake twice.

How do you structure the negatives?

Build the negatives so they scale, with two kinds of list. One is a permanent, account-wide list of terms that never apply anywhere you advertise, the job-seeker and free-tool queries, added once and reused on every campaign. The other is campaign-specific, for terms that are wrong here but fine elsewhere, like a beginner query under a campaign built for enterprise buyers. Match type matters too: a phrase-match negative blocks a whole span of queries, while an exact-match negative blocks only the one, so reach for phrase when a theme is off and exact when a single query slips through.

Check a negative before you save it. It is easy to block a term so broad that it takes real queries down with it, so read what each one would catch across the account, not just the phrase in front of you.

The search terms report rewards attention and punishes neglect. Reviewed weekly, it is one of the highest-return habits in a paid account. Programmatic CMO's Google Ads agent reads it every morning, flags the terms spending without converting, and proposes negatives for your approval, so the cull runs continuously instead of once a quarter.