Google Ads Search Terms Report Best Practices

The best Google Ads Search Terms Report practices make actual searches useful for a specific account decision. Choose the right scope and date range, keep the matching keyword and match type visible, compare spend with qualified outcomes, classify intent before acting, and use Search Terms Insights to understand grouped demand without pretending that every query is exposed. A report review is successful when it produces a small, defensible action or a clear decision to wait.

The report shows searches that triggered ads, not the full universe of searches. Google omits some low-activity terms for privacy, while Search Terms Insights groups additional demand into themes and subthemes. Google’s current Search Terms Report guidance explains those surfaces and their differences. This page focuses on operating practices; the Search Terms Report settings guide covers the basic view configuration.

Start with a question and an owner

Do not open the report with the vague goal of “finding wasted spend.” Decide whether the review is about irrelevant queries, new demand, ad and landing-page language, a match-type pattern, a location or device issue, brand coverage, or conversion quality. The question determines the scope, columns, date range, and action.

Assign an owner who can apply or approve the resulting change. A query review may produce a negative keyword, a new keyword, a new ad group, a landing-page revision, a conversion repair, a budget decision, or no change. Record the decision type before sorting rows so a report does not become an unowned export.

Match the review to the business outcome

Use the outcome the campaign is meant to produce. For lead generation, compare forms with qualified stages, accepted leads, pipeline, or revenue when those imports are reliable. For ecommerce, include purchase value, returns, margin, and customer type where available. For local services, check service-area fit, call quality, and scheduling capacity. For awareness or research, define the intended reach or engagement outcome instead of pretending every click is a sale.

Keep click-through rate, cost per click, and raw conversions as diagnostic measures. They can help prioritize a row, but they should not decide whether a query is valuable when the business outcome is different.

Choose the report surface and date range deliberately

Use the regular Search Terms Report when you need individual exposed searches, the matching keyword, match type, campaign, ad group, and outcome. Use Search Terms Insights when you need grouped categories or subthemes that include demand not shown as individual rows. The two surfaces answer related but different questions and should not be expected to reconcile exactly.

Choose a date range that includes the campaign’s normal conversion lag and enough spend or impressions to support a decision. Avoid using today or a partially processed period for a final conclusion. Annotate promotions, budget shifts, bid changes, tracking releases, website outages, seasonality, and policy changes.

Keep a stable comparison window

When reviewing a change, save the before-state and the date the change took effect. Compare a stable period before and after while accounting for conversion processing and other changes. If several settings changed together, report the result as uncertain rather than assigning the whole effect to the Search Terms review.

Do not use an arbitrary seven-day rule for every account. A high-volume ecommerce campaign and a low-volume B2B campaign need different evidence. Choose a window that reflects the business’s sales cycle, conversion lag, and ability to respond to wasted spend.

Keep the fields that explain a decision visible

At minimum, keep Search term, Keyword, Match type, Campaign, Ad group, impressions, clicks, cost, conversions, conversion rate, conversion value, and the efficiency metric that matters. Add qualified outcomes, revenue, margin, device, location, network, or landing-page information when they answer the review question.

The Search term is what the person entered. The Keyword shows which account keyword matched it. Match type describes the relationship between the search and the keyword; do not assume it is simply a copy of the keyword’s configured match setting. Campaign and Ad group show where the decision belongs.

Save a named column set and date-range convention after testing it. A repeatable view makes reviews comparable and reduces the chance that one analyst sorts by clicks while another sorts by cost and both call the result the same kind of audit.

Classify intent before choosing an action

Read each important search as a customer question. Compare it with the keyword, match type, ad group, ad, landing page, location, device, and business outcome. Then classify it as one of four working states:

Valuable and relevant: it fits the offer and produces, or has a credible path to produce, the desired outcome.
Relevant but unproven: it fits, but the query has not accumulated enough evidence for an economic decision.
Irrelevant or unserviceable: it asks for something the business does not sell, cannot deliver, or should not pay to attract.
Ambiguous: the wording could mean several things and needs more context, volume, or a page review.

This classification prevents the common mistake of treating every query without a conversion as waste. A long sales cycle or low-volume row may need more time. It also prevents a high-click query from being kept merely because it is active when the business cannot serve it.

Use a decision record

For a meaningful row or theme, record the search, intent, matching keyword, scope, spend, outcome, decision, reason, owner, and date. Choose one action: retain, add coverage, add a negative, revise an ad, improve the landing page, investigate tracking, restructure, or wait. A “wait” decision should include the evidence threshold and next review date.

Add and remove coverage with the narrowest safe rule

When a search represents valuable demand, check whether an existing keyword, ad group, ad, and page already cover it. If they do, improve the message or destination rather than adding a duplicate keyword. If it represents a distinct product, service, customer problem, location, or stage, create coverage where the budget and policy can answer that intent.

When a search is clearly irrelevant or unserviceable, add a negative keyword at the narrowest scope that solves the problem. Campaign-level negatives are appropriate when the rule applies to every ad group; ad-group or shared controls need a deliberate scope review. Test the exclusion against valuable searches you want to keep, including spelling, plural, location, brand, and qualifier variants.

Do not turn every successful search into an exact keyword. A new keyword can duplicate existing coverage, fragment data, or change the match behavior without improving control. Choose the structure and match type based on the business question, expected demand, and need for isolation.

Review the effect of a negative

After an exclusion, check eligible impressions, spend, conversions, query themes, and value for unintended loss. A negative that lowers cost can still remove high-value demand. Record the scope, expected effect, and rollback condition so the rule does not become permanent by inertia.

Use Search Terms Insights without overclaiming

Search Terms Insights groups search demand into categories and subthemes, including some low-volume activity that is not exposed as an individual row. Use it to understand customer language, product and service categories, emerging themes, and coverage gaps. Use the regular report when you need the exposed search, matching keyword, and direct account action.

Do not expect the two surfaces to reconcile exactly. Google notes that privacy aggregation, hidden low-activity queries, conversion processing, and theme-label history can create differences. Search Terms Insights labels can be generated from a recent history window and may include terms that are no longer serving or have since been excluded. Read the date range and surface before turning a theme into a keyword or negative.

Treat hidden queries as a category signal

When a query is omitted for privacy, use the grouped theme and overall campaign pattern instead of trying to reconstruct the individual search. Do not claim that an unseen query did not occur or that a single visible row represents all demand in that category. The absence of a row is a reporting limitation, not proof of zero traffic.

Connect queries to ads and landing pages

Search terms are customer language, but they are not automatically good ad copy. Repeated relevant phrasing can improve a headline, description, sitelink, landing-page section, or product taxonomy when it is accurate and natural. Check that the page offers the same product, service, location, qualification, price context, and next step.

A high-click query with a low qualified rate may reveal a message or page mismatch rather than a keyword problem. Compare query, ad, page, device, location, conversion, and business quality before adding a negative. Conversely, a query that fits the offer but converts poorly because the page is broken should trigger a page or tracking repair.

Keep legal, brand, privacy, and customer information out of casual exports and shared documents. Query data can contain sensitive intent. Limit access, use the account’s reporting controls, and retain only the data the review needs.

Review on a cadence the account can use

Set a cadence based on volume, conversion lag, policy risk, and the speed at which the team can apply a decision. High-spend accounts may need a regular review; low-volume accounts may need a less frequent but deeper review. An automated schedule is not useful if no owner can approve negatives, landing-page changes, or conversion repairs.

At each review, compare:

query and theme relevance;
matching keyword and match-type patterns;
spend, conversions, value, qualified outcomes, and lag;
ad and landing-page alignment;
device, location, language, schedule, and network differences;
changes to budget, bidding, tracking, pages, offers, and policy.

Use a stable saved view, consistent date conventions, and a short decision log. If the account has Search partners or another network in scope, inspect that segment separately when the search format or quality differs.

Escalate uncertainty instead of guessing

Escalate a query when the result could affect brand policy, regulated terms, customer privacy, a universal negative, a large budget, or a material landing-page change. An ambiguous query is often a request for more context, not permission to block or expand it.

Common Search Terms Report mistakes

Avoid these patterns:

reviewing today’s partial data as if it were a final performance period;
sorting by clicks and calling the highest rows “best” without qualified outcomes;
treating a hidden query as evidence that no demand exists;
expecting Search Terms Insights and the regular report to match row for row;
adding every successful query as a duplicate keyword;
adding a broad negative because one variation looked irrelevant;
changing match types, negatives, ads, pages, budget, and bids together;
ignoring the matching keyword, ad group, device, location, or landing page;
exporting sensitive query data without an access and retention policy;
treating a report export as an action plan with no owner or review date.

Search Terms Report best-practice checklist

Before closing the review, confirm:

[ ] The question, scope, owner, date range, conversion lag, and business outcome are documented.
[ ] Search term, Keyword, Match type, Campaign, Ad group, cost, conversions, value, and quality fields are available.
[ ] Queries are classified by relevance and evidence before an account change is selected.
[ ] Negative keywords and new coverage use the narrowest safe scope and are tested against wanted variants.
[ ] Search Terms Insights is used for grouped demand and the regular report for exposed query decisions.
[ ] Privacy limits, hidden queries, conversion lag, and theme-label history are understood.
[ ] Ads, landing pages, devices, locations, networks, tracking, and business quality are reviewed with the query.
[ ] Every action or wait decision has an owner, date, rationale, and next review.

For configuration details, see the best Search Terms Report settings guide. For applying query findings, see how to optimize the Search Terms Report. The adz.so product overview describes the current audit entry point and does not imply that adz.so automatically manages Search Terms Report decisions.