The Google Ads Search Terms report turns real queries into practical campaign decisions. This workflow shows how to classify search terms, estimate their value, add promising queries as keywords, build a useful negative keyword list, and decide when a query deserves a new ad group, landing page, or campaign test.
Overview
A Google Ads keyword strategy should not rely only on forecasts, seed terms, or a keyword research tool. The Search Terms report shows what people actually searched before clicking an ad. That makes it one of the most useful inputs for PPC keyword research and intent mapping.
Use the report to answer five questions:
- Which searches produced valuable conversions?
- Which queries reveal a stronger or weaker intent than the current keyword suggests?
- Which terms are consuming spend without supporting the campaign goal?
- Which recurring themes should become new Google Ads keywords or dedicated ad groups?
- Which searches should be excluded, and at what level?
Do not treat every unusual query as a problem. A query can be relevant even when it has not converted yet, particularly when it belongs to a high-intent theme with limited data. The goal is to make decisions using relevance, intent, cost, conversion evidence, and landing-page fit together.
For the review itself, create a working sheet with these columns: date range, campaign, ad group, search term, matched keyword, match type, intent, conversions, conversion value, cost, clicks, landing page, action, reason, owner, and review date. Keeping the reason and owner visible prevents the same ambiguous query from being debated repeatedly.
How to estimate
Start by deciding what “valuable” means for the campaign. A purchase campaign may prioritize conversion value and return on ad spend. A lead-generation campaign may use qualified leads or estimated pipeline value. A SaaS campaign may need to separate free trials, product-qualified accounts, and sales opportunities rather than counting every form submission equally.
Use a simple value estimate for each search-term group:
Estimated conversion value = conversions × assigned value per conversion
Then compare that estimate with spend:
Estimated return ratio = estimated conversion value ÷ cost
If a term has no recorded conversion, do not automatically assign it a value of zero. Instead, classify its evidence level:
- Proven: the term has enough conversion evidence to support a direct action.
- Promising: the term is closely relevant, shows commercial intent, or has useful engagement but insufficient conversion data.
- Unclear: the query could be relevant, but its intent or business value is not yet established.
- Wasteful: the query is irrelevant, unsuitable for the offer, or conflicts with the campaign objective.
Next, assign an action. A practical decision set is:
- Promote: add the query or its normalized theme as a keyword when it is relevant and supported by performance.
- Observe: retain it for another review when the theme is relevant but the sample is too small.
- Exclude: add a negative keyword when the query is clearly outside the offer, audience, geography, or buying stage.
- Restructure: move the theme into a more appropriate campaign or ad group when the current message, budget, bid strategy, or landing page does not fit.
- Investigate: check conversion tracking, search intent, and landing-page behavior before changing targeting.
When you promote a query, avoid creating near-duplicate keywords without a reason. Record why the new keyword exists, which ad group owns it, and what match behavior you expect. The same discipline makes keyword grouping for PPC easier to maintain across Google Ads and Microsoft Ads.
Inputs and assumptions
Search-term analysis is only as reliable as its inputs. Before making changes, verify the following:
Conversion definition
Confirm that the conversion column represents a meaningful business action. If micro-conversions and primary conversions are mixed together, separate them before estimating value. A tracked page view should not be compared directly with a qualified sale unless the values are intentionally modeled.
Cost and date range
Use a date range long enough to show a useful pattern, but short enough to reflect the current offer, budget, and landing page. A weekly review can identify emerging queries; a longer monthly review is better for judging conversion efficiency. Note major changes such as a new bid strategy, pricing page, promotion, or tracking setup.
Intent labels
Use labels that match the business. A practical set is informational, commercial investigation, transactional, navigational, competitor, support, employment, and irrelevant. Add a confidence field when intent is debatable. For example, “accounting software for freelancers” is usually more commercially useful than “what is accounting,” but the final classification still depends on the offer and landing page.
Landing-page fit
Compare the query with the ad and the destination page. A relevant term can perform poorly when the page does not answer the specific need. Review message match before excluding a query that appears commercially relevant. The landing page message match audit provides a separate checklist for this decision.
Privacy and incomplete query data
Some searches may not appear in the report or may be grouped because the available data is limited. Treat the visible report as an important sample, not a complete record of every search. Avoid making broad conclusions from a small set of exposed queries.
For consistent reporting, keep campaign names, landing pages, and conversion actions documented. A reliable conversion tracking audit should precede major keyword decisions when the numbers appear inconsistent.
Worked examples
Example 1: Promote a high-intent theme
A software campaign sees several searches containing “project management software for agencies.” The queries are tightly aligned with the product, reach the intended landing page, and produce two qualified leads during the review period. The campaign spent $180 on that theme, and the team assigns an internal value of $250 to each qualified lead.
Estimated conversion value: 2 × $250 = $500. Estimated return ratio: $500 ÷ $180 = 2.78.
The correct action is not simply to copy every visible variation. Group the recurring theme, check whether the existing ad speaks specifically to agencies, and consider promoting a carefully chosen keyword into an ad group with a tailored message. If the landing page remains general, the right action may be a landing-page test rather than a keyword expansion.
Example 2: Exclude a mismatched intent
An ecommerce campaign selling paid accounting software receives repeated searches for “free accounting spreadsheet template.” The query may be related to accounting, but it signals a different solution and likely a different expectation. If the business does not offer a free template, classify the term as wasteful and add an appropriate negative keyword after checking for legitimate product language that might be unintentionally blocked.
Document the reason as “free resource intent; no matching offer.” This is more useful than recording only “bad query,” especially when another team reviews the negative keyword list later.
Example 3: Restructure an emerging intent
A B2B campaign targets a broad category, but the report increasingly shows searches from a specific industry. The terms have clicks and early engagement but not enough conversions to prove efficiency. Rather than excluding them, label the theme as promising. Create a controlled test with an industry-specific ad, a relevant page, and a defined budget or bid constraint. Revisit it after enough exposure to compare against the original broad group.
This approach separates discovery from permanent account structure. It also prevents the common mistake of allowing a new intent to alter the message of an established ad group without measurement.
When to recalculate
Use a weekly search-term review for active accounts, but reserve structural changes for patterns rather than isolated queries. At minimum, revisit the analysis when the following inputs change:
- Budget, bids, bidding strategy, or campaign goals change.
- Pricing, promotions, product packaging, or availability changes.
- A new landing page, offer, or conversion action is launched.
- Tracking is repaired, replaced, or found to be unreliable.
- A competitor, industry term, or seasonal need begins appearing repeatedly.
- Search behavior shifts enough to create a new intent or audience segment.
Use this repeatable rhythm:
- Weekly: scan new queries, tag obvious intent, catch urgent waste, and assign unresolved terms.
- Monthly: calculate value and cost by theme, promote proven opportunities, review negative keyword coverage, and inspect landing-page fit.
- After a major change: reset the comparison period so old performance does not obscure the new offer or targeting setup.
- Twice yearly: review the account’s intent taxonomy, campaign structure, and recurring themes for outdated assumptions.
End every review with three decisions: what to add, what to exclude, and what to test next. Save the report with its date range and assumptions, then assign an owner and a follow-up date. Pair query findings with your negative keyword workflow, document tracking with a consistent UTM naming convention, and compare paid-search themes with organic coverage through SEO and PPC keyword overlap analysis. The result is not a one-time cleanup: it is a durable system for turning changing search behavior into measured keyword decisions.