A search-term report is useful only when it leads to a decision. This workflow shows how to classify paid search queries, estimate their value, and turn the results into keyword, negative keyword, ad copy, landing-page, and budget actions. It also includes a repeatable worksheet and scoring method for weekly or fortnightly reviews.
Overview
Search-term analysis is the process of reviewing the actual queries that triggered your ads rather than relying only on the keywords you added to an account. The report can reveal search language you did not anticipate, irrelevant variations, new commercial themes, and gaps between an ad promise and the landing page.
The goal is not to change every keyword after every export. A useful review separates queries into four broad groups:
- Keep or monitor: Relevant queries that fit the offer and have acceptable performance or useful learning value.
- Promote: Relevant queries that deserve a dedicated keyword, ad group, landing page, or bid treatment.
- Exclude: Queries that are outside the offer, audience, geography, or buying intent. These may require a negative keyword, subject to checking how the exclusion will affect related searches.
- Investigate: Queries with unclear intent, limited data, tracking concerns, or conflicting signals.
Use the same definitions across review periods. Consistency makes it easier to compare recurring themes and distinguish a one-off query from a genuine keyword opportunity. For account structure, document any new themes using the naming rules in Keyword Grouping for PPC Campaigns.
How to estimate
A simple estimation model helps prioritize work when a report contains hundreds or thousands of rows. Start by assigning each query an action score rather than treating every row as equally important.
For each query, record:
- Relevance: Score from 0 to 3. Use 0 for unrelated, 1 for loosely related, 2 for relevant, and 3 for a close match to the offer.
- Intent: Score from 0 to 3. Use a higher score when the wording indicates a stronger likelihood of taking the desired action. Informational, navigational, comparison, and transaction language can be used as practical categories.
- Conversion evidence: Record conversions, conversion value, or another agreed business outcome. Do not combine incompatible conversion actions without labeling them.
- Cost exposure: Record spend and clicks for the selected date range. A query with little spend may need monitoring, while a query with substantial spend and no useful outcome deserves faster attention.
- Action confidence: Score from 1 to 3. A score of 3 means the next action is clear; 1 means the data or intent needs more investigation.
One practical priority formula is:
Priority score = cost exposure × action confidence × business relevance.
For a lightweight version, use cost exposure as low, medium, or high; business relevance as low, medium, or high; and confidence as low, medium, or high. If you prefer numbers, multiply the three values after assigning each a 1–3 score. This is a sorting aid, not a forecast. It prevents a large report from being reviewed in an arbitrary order.
For queries with conversion data, add an outcome check. Compare cost per conversion or return on ad spend with the target used for that campaign. If a query has no conversion yet, do not automatically classify it as poor. Consider its spend, click volume, sales cycle, assisted role, and tracking quality before acting.
Inputs and assumptions
Export a consistent set of fields from each advertising platform. Labels vary, but a useful PPC search query report usually includes the query, campaign, ad group, matched keyword or targeting element, match type where available, impressions, clicks, spend, conversions, conversion value, and date range. Add device, location, network, or audience fields when they affect the decision.
Build a worksheet with these columns:
- Query and normalized query
- Campaign, ad group, and matched keyword
- Intent category
- Relevance score
- Impressions, clicks, cost, conversions, and value
- Landing page and message-match note
- Recommended action
- Owner, date, and completion status
Normalize carefully. Lowercase text for comparison, remove accidental spacing, and preserve the original query in a separate column. Avoid merging terms that differ in meaning, such as a product name and a support-related phrase. Also separate close variants when geography, service level, product specification, or audience changes the commercial intent.
Use conversion data only after checking the conversion setup and attribution window used in the report. A conversion column may include more than one action, and platform-reported outcomes may not match analytics or CRM records. For a broader check, use the PPC Campaign Audit Checklist and document which source is used for final business decisions.
Keep negative-keyword decisions focused here on classification and prioritization. Before applying exclusions, check campaign scope, close variants, and possible interactions with other ad groups. The detailed implementation and audit process belongs in the Negative Keyword Management Playbook.
Worked examples
Example 1: Local service campaign
A local service advertiser finds a query containing the service name and the target city. It has several clicks, one tracked lead, and a landing page that directly describes the service. Relevance is 3, intent is 3, and action confidence is 3. The recommended action is to keep monitoring and consider promoting the query to a tightly named keyword group if the pattern recurs. Before doing so, confirm that the new structure will not create unnecessary overlap.
A second query asks about training or certification rather than hiring the service. It has spent a meaningful amount without a lead. Relevance is 1, intent is 1, and cost exposure is high. Classify it as an exclusion candidate, then check whether a phrase-level or exact exclusion better protects relevant searches.
Example 2: Ecommerce campaign
An ecommerce query includes a specific product model and a purchase-oriented phrase. It has strong relevance but no conversion in a short date range. Instead of excluding it, mark it for investigation. Check stock availability, price or offer alignment, product-page load experience, and conversion tracking. If similar queries recur, create a product-focused keyword group or ad variation and link it to the most relevant page.
Example 3: B2B campaign
A B2B query matches the company’s solution but uses early research language. It produces engaged visits but no form submission. Do not force a direct-response judgment if the campaign also supports demand creation. Mark the query as relevant, record its funnel stage, and compare it with CRM or assisted-conversion information where available. The appropriate action may be a content or message test rather than a negative keyword.
When a query repeatedly exposes a message mismatch, review the ad and destination together. The Ad-to-Landing-Page Message Match Audit provides a useful companion checklist. If query themes are being compared across platforms, keep campaign names, UTM values, and reporting dimensions consistent; see the UTM Parameter Guide for Paid Ads.
When to recalculate
Run the workflow weekly for accounts with enough activity to produce meaningful changes, or fortnightly when volume is lower. Recalculate priorities after a major budget change, landing-page update, targeting change, conversion-tracking change, product launch, seasonal shift, or bidding strategy test. A query that was too small to judge last period may become actionable after additional clicks or conversions.
At each review, compare the new export with the previous worksheet. Look for recurring query themes, repeated irrelevant wording, newly emerging products or services, and actions that were recommended but not completed. Record the reason for every material change so later performance can be interpreted in context. A bid or budget change should be logged separately using a change record such as the PPC Bid Strategy Change Log.
Refresh the worksheet when platform export fields, conversion definitions, attribution settings, or reporting filters change. Recheck your assumptions whenever business goals change: lead quality may matter more than lead volume, profit may replace revenue as the optimization target, or a new service area may alter what counts as relevant. Finish each review with three concrete outputs: the highest-priority query actions, the person responsible for each action, and the date when results will be checked. That turns search-term analysis from a report-reading exercise into a repeatable keyword optimization system.