Google Ads Search Query Mining Guide: How to Find Winning Keywords and Cut Wasted Spend

This google ads search query mining guide gives you a repeatable routine for turning the Search Terms Report into new keywords, negative keywords, and a cleaner campaign structure. You'll work through one loop: review, bucket, add, negate, cluster, measure. Before you start, make sure you have a Google Ads account with at least a few weeks of Search campaign data and permission to edit keywords and negatives. Nothing here needs a spreadsheet, though you may want a notes file for the change log in the last step.

Step 1: Set Up Your Search Terms Report and Date Range

First, the vocabulary. A keyword is what you bid on. A search term is what the person actually typed. Mining means moving good search terms into your keyword list and bad ones into your negative keywords, which tell Google what not to show your ads for. If the distinction still feels fuzzy, this breakdown of search terms vs keywords in Google Ads covers it in more depth.

As of November 2026, you'll find the report in Google Ads under Insights and reports, then Search terms. Google renames menus fairly often, so if you don't see it there, check the current path in Google Ads Help.

Next, pick a date range that matches your volume. This is illustrative guidance, not a Google rule:

• Low-traffic accounts: roughly 30 to 90 days, so you have enough queries to see patterns.
• High-spend accounts: shorter windows, such as the last 7 to 30 days, so you aren't re-reviewing terms you already handled.

Then add the columns that let you judge a term without guessing: cost, clicks, conversions, conversion value, CTR, and the Added/Excluded status. Filter out terms already added as keywords or excluded as negatives so you only see unreviewed queries.

One limit to know before you start. Google only lists search terms that meet its query-privacy thresholds, so some of your spend sits in a row called "Other search terms." You can't mine what you can't see, so don't assume the visible list covers every dollar. If that row is large, pay extra attention to your match types and your negative lists, because they're your only levers over the hidden queries.

Step 2: Sort and Filter to Surface the Terms Worth Reviewing

Run two passes over the same data rather than scrolling top to bottom.

1. Spenders first. Sort by cost, descending. You're looking for terms that burned budget with no conversions.
2. Winners second. Sort by conversions, descending. You're looking for terms that earn their keep and may not be keywords yet.

Filters speed this up. Try Conversions = 0 combined with a cost threshold to isolate expensive non-performers. For winners, try Conversions greater than 0 and narrow to terms triggered by broad or phrase match keywords, since those are the ones most likely to be unclaimed. If you want more filter ideas, see this walkthrough on how to filter search terms in Google Ads.

Check the match type column, labelled "Search term match type" in the report. It shows how the query matched, and pairing it with the keyword that triggered it tells you which of your keywords is pulling in what. A loose broad match keyword producing five irrelevant queries is a different fix from five junk queries spread across your whole account.

Set a threshold before you judge

A term with 3 clicks and no conversions isn't proof of junk. Decide on a minimum before you start, such as a click count or a spend level relative to your typical cost per conversion. Anything under the line goes on a watch list instead of the negative list. The exact number depends on your account, so treat it as your own rule rather than a standard.

The most common mistake here is reviewing only the top 20 rows. Wasted spend often lives in the long tail: hundreds of small queries that each cost little but share the same bad intent. Scroll deeper, or sort alphabetically once, so repeating patterns jump out. If the volume feels unmanageable, this guide to reviewing the Google Ads search terms report faster can help. Step 6 deals with those patterns directly.

Step 3: Sort Search Terms Into Four Buckets

Every reviewed term goes into one of four buckets. Deciding the bucket first keeps you from making changes in the middle of reading.

1. Add as keyword. Relevant, converting or clearly high-intent terms that aren't in the account yet.
2. Add as negative. Clearly irrelevant queries: job seekers, "free," DIY, the wrong product.
3. Watch. Relevant but unproven. Leave them alone until they have more data.
4. Restructure. Relevant terms that are triggering the wrong ad group or campaign and belong somewhere else.

Here's an illustrative example for a plumbing campaign. "Emergency plumber near me" is relevant, urgent, and likely to convert, so it becomes a keyword. "Plumber salary" is someone researching a career, so it becomes a negative. "Water heater repair cost" is relevant but you haven't seen enough data, so it goes on the watch list. "Drain cleaning" showing up in your general plumbing ad group when you have a dedicated drain campaign goes in the restructure bucket.

Notice that performance alone doesn't decide the bucket. A term with zero conversions might still be a perfect prospect that just hasn't had enough clicks. A term with one lucky conversion might be a fluke from someone who can never become a repeat customer. The question to ask is whether this searcher could ever become your customer. If the honest answer is no, it's a negative regardless of how the numbers look. If the answer is yes, give it the data it needs. For help spotting the "no" cases, see how to find irrelevant search terms in Google Ads.

Keep the watch bucket short and revisit it each cycle. If it grows without ever being emptied, your threshold from Step 2 is probably too high.

Step 4: Add High-Intent Search Terms as Keywords with the Right Match Type

Match type should follow the purpose of the keyword:

• Exact match for proven converters, where you want tight control over what triggers the ad.
• Phrase match for controlled expansion around a theme that works.
• Broad match only when you have solid conversion tracking and use smart bidding, since broad leans on those signals to filter intent.

Google has changed how match types treat close variants and meaning over the years, so verify the current behavior in Google Ads Help before you rely on any assumption about what a phrase or broad keyword will trigger as of November 2026.

Place and group before you add

Put each term in the most relevant ad group, so the ad copy and landing page match what the person typed. A query about "emergency plumber" should land in the ad group with emergency messaging, not the general one.

Check for cannibalization too. If an existing keyword in another ad group already covers the term, adding it again splits data and makes it harder to see which one performs. Search your keyword list for the term before you add it.

Resist building one ad group per search term. Group close variants into themes, so "emergency plumber," "24 hour plumber," and "urgent plumber near me" live together with one set of ads. Hundreds of single-term ad groups become unmanageable fast. If you need a refresher on the mechanics, here's a step-by-step guide on how to add keywords to Google Ads.

Doing it inside the report

This is where Keywordme fits. It's a Chrome extension that works inside the Search Terms Report, so you can select terms, apply a match type, and add them as keywords without leaving Google Ads or exporting anything. You still review each term and choose the match type and destination yourself; the extension speeds up the clicking, not the judgment.

Step 5: Add Negative Keywords at the Right Level and Match Type

Start with the level, because a negative in the wrong place either does nothing or blocks too much.

• Ad group: for conflicts between ad groups, such as keeping "repair" queries out of your "installation" group.
• Campaign: for junk specific to that campaign.
• Shared negative keyword list: for themes common across campaigns, applied in one go.

Google also offers account-level negatives. Confirm current availability and list limits in Google Ads Help, since both have changed over time. For the full process, see this guide to adding negative keywords in a Google Ads campaign.

Match types work differently for negatives

Negative keywords don't expand to close variants or synonyms the way positive keywords do. Block "job" and you may still see "jobs" or a misspelling, so add plurals and common misspellings deliberately. Verify the current rules in Google Ads Help before you build a large list.

For choosing the match type, a simple split works:

• Single-word negatives for universal junk, like "jobs" or "free."
• Phrase or exact negatives for specific conflicts, where the same word is fine in other contexts.

Watch for the broad negative that blocks good traffic. If your offer is budget pricing, adding "cheap" as a negative could cut off your best customers. When a word is ambiguous, use a phrase or exact negative for the specific bad query instead.

Agencies get real value from a shared list. Keep one for themes that recur across clients' accounts, like job and career terms, and apply it to each new account as part of setup.

In the Search Terms Report, Keywordme lets you mark junk terms and build negative lists with a click, or in bulk. As an illustration, you might select ten "free" and "jobs" queries at once and send them to a shared list. You still decide the level and match type for each batch, and you should read the selection before you confirm it. If your report is full of clutter, this walkthrough on removing junk search terms in Google Ads goes further.

Step 6: Look for Patterns and Cluster Recurring Words

Term-by-term cleanup is slow. Once you've done a pass, look at the words that repeat across many low-value queries. If "free," "template," or "reddit" shows up in dozens of queries, block the pattern once instead of adding each full query.

The same scan works for good signals. Repeated location names, product model numbers, "near me" phrasing, or problem-based wording ("leaking pipe under sink") can point to a new ad group or a dedicated landing page.

How to find the patterns

This is word-level analysis, sometimes called n-gram analysis: breaking queries into single words or short word groups and totaling cost and conversions for each. You can do it manually by sorting the report and grouping similar queries, or by exporting and using a pivot table. Keywordme also has a keyword clustering feature that groups related terms inside the report. Check the current feature details on the Keywordme site before relying on specifics, as they may change. For more on analysis methods, see these Google Ads search term analysis tips.

Brand and competitor queries

While you're scanning, look for queries containing competitor names or your own brand. Decide on each separately. Competitor terms may be worth targeting in a dedicated campaign or worth excluding, depending on your margins and goals. Your own brand queries usually perform very differently from non-brand ones and are better kept apart, so reports aren't skewed.

The key mistake to avoid: negating a word that appears in both bad and good queries. Before you block "cheap" or "template," filter the report for that word and check the converting queries first. If it shows up in winners, use phrase or exact negatives for the specific bad combinations instead.

Step 7: Build a Repeatable Review Routine and Measure the Impact

Mining works as a habit, not a cleanup project. Set a cadence by spend. As a rough, illustrative guide, large budgets benefit from a weekly review, while smaller accounts can go every two weeks to monthly. Fast-moving accounts accumulate junk queries quicker, so the cadence should follow how quickly new terms appear, not the calendar. A documented search terms report workflow makes the cadence easier to keep.

Keep a simple change log with these fields:

• Date of the review
• Terms added as keywords
• Negatives added
• Level and match type for each change

The log lets you trace a performance shift back to a specific change, and it stops the next reviewer from undoing your work.

Measure after two to four weeks. Compare these on the affected campaigns against the period before your changes:

• Share of spend going to non-converting or irrelevant terms
• Conversion rate
• Cost per acquisition (CPA)
• Search impression share

Be careful with attribution. Bidding strategy changes, seasonality, new ads, and budget shifts can all move these numbers, so note anything else that changed in the same window. If you changed several things at once, you can't credit the negatives.

For agencies, turn the routine into a checklist you run for every client: date range, filters, four buckets, adds, negatives, shared list updates, log entry. Standardizing it makes reviews faster and easier to hand off.

Set honest expectations too. Mining tends to cut waste and improve relevance, but results vary by account, and nothing here is guaranteed.

Run Your First Mining Pass on Your Biggest Campaign

Start with your highest-spend campaign. Go through the steps once, leave the changes alone for a few weeks, then check the results against your log. After that, put the review on your calendar so it becomes routine.

If the clicking is what slows you down, Optimize Google Ads Campaigns 10X Faster. Without Leaving Your Account. Keywordme lets you remove junk search terms, build high-intent keyword lists, and apply match types instantly, right inside Google Ads. No spreadsheets, no switching tabs, just quick, seamless optimization. Start your free 7-day trial (then just $12/month) and take your Google Ads game to the next level.