Google Ads Search Terms Review Checklist in 15 Minutes

Once a week, sort the last 7 days of search terms by cost. Run each term that spent money through five questions: did it convert in the last 90 days, could this searcher ever buy, has it spent enough to judge, is the problem the term or your landing page, and does it deserve its own keyword? Every term ends up as block, keep, watch or add as keyword. Nothing gets left as "not sure".
Most of the 15 minutes goes on the borderline terms: right intent but no conversions yet, a single low-value conversion, competitor names, research questions. Before you block one of these, wait until it has spent about one target cost per conversion. When you do block, use the narrowest match type that covers the problem, and keep a watch list so next week starts where this week stopped.
What a weekly search terms review is for
A weekly search terms review has one job: stop paying for the same bad search term twice. It is not a full account audit. It is a short, repeatable pass over last week's new spend. Each search term gets one of four outcomes: block, keep, watch or add as keyword.
Most checklists fail because they only cover the easy cases. Everyone knows to block "plumber salary" when they sell emergency plumbing. The hard part is "plumber cost to replace water heater". It has spent $60, has no conversions and sounds like a buyer. If your checklist has no rule for that term, you decide it differently each week. Some weeks you block a term that would have converted next month. Other weeks you keep paying for one that never will.
The routine below takes about 15 minutes for an account that spends a few thousand dollars a week, once the view is set up. Larger accounts need more time or more automation. That is not because the steps change. There are simply more terms above the cost line. If you need a deeper one-off pass, for example when you take over a new account, the 20-minute search terms analysis is the better starting point. This checklist is for the weeks after that.
Set up the view once, then reuse it
Save one search terms view with the right date range, columns and sort. Then the weekly review starts with judging, not with clicking through menus.
Here is what the saved view needs:
- Date range: last 7 days, compared with nothing. You are looking for new spend. You are not looking for trends.
- Columns: search term, match type, campaign, ad group, keyword, clicks, cost, conversions, conversion value (if you track it) and cost per conversion.
- Sort: cost, descending.
- Filter: cost greater than a floor. Set the floor at roughly a tenth of your target cost per conversion. With an $80 target, filter to terms that cost more than $8 this week. The rest still matter in total. They are a monthly job, not a weekly one, and low-volume terms add up in a way that needs its own pass.
The search terms report filters that surface waste are covered in detail elsewhere. For this routine you only need the cost sort and the floor.
You also need one more thing: a watch list. A plain spreadsheet is enough, with four columns: the term, the date you first flagged it, its cost so far and what you are waiting for. Without a watch list, every borderline term gets judged from scratch each week, which takes longer and gives worse decisions.
The 15-minute checklist
Each block of time has one task. Do them in order and stop when the timer runs out. Whatever is left over goes on the watch list.
Minutes 0–2: check the watch list first. Open last week's list. For each term, check its cost and conversions since you flagged it. Some will now have enough data to decide. Decide them now, using the tree in the next section. Terms that still lack data stay on the list.
Minutes 2–4: scan the top 10 by cost. These terms carry most of last week's spend, so read every one of them carefully. A single bad term in the top 10 can cost more than everything below it.
Minutes 4–10: run the decision tree on every term above the cost floor. Most terms take seconds. "plumbing apprenticeship near me" is a block. "emergency plumber [your city]" with two conversions is a keep. Only the borderline terms take real thought, and the tree exists so that even those take under a minute.
Minutes 10–13: apply the decisions. Add the negatives at the right level with the right match type (more on that below). Add any new keywords. Move any watch terms into the sheet.
Minutes 13–15: look for a pattern. Read your new negatives as a group. Five blocked terms that all contain "jobs", "salary" or "hiring" are not five problems. They are one missing negative that should have been there from launch. Two blocked terms naming towns you don't serve point to a location setting problem, not just a search terms problem. Two minutes of pattern-reading here saves you from blocking the same idea word by word for months.
The decision tree: block, keep, watch or add as keyword
Ask the five questions in this order and stop at the first one that gives an answer. The order matters. It puts the costliest mistake first, which is blocking a term that converts.
1. Did it convert in the last 90 days?
If yes, keep it. Stop here. A term with a conversion is not a candidate for blocking this week, however expensive it looks. It may still be a bad term, for example one that converts at a terrible cost. That is a question about bids and value, and it belongs in question 5 or in a separate review. It does not belong in a negative keyword list.
Look at 90 days, not the 7 days on screen. Conversions are credited to the date of the click, and many accounts have a lag of days or weeks between click and conversion. A term that looks dead this week may already have a conversion recorded from last month.
2. Could this searcher ever buy what you sell?
If the answer is clearly no, block it. These are the easy cases:
- Job seekers: "plumber salary", "plumbing apprenticeship", "hiring plumbers"
- Do-it-yourself with no buying intent: "how to unclog a toilet without a plunger"
- Wrong product: "plumbing supplies wholesale" when you sell a service
- Free or education: "free plumbing course", "plumbing code pdf"
- Places you don't serve
No amount of data will turn these into buyers, so you don't need to wait for spend. Block on the first click.
If the answer is "maybe" or "yes", go to question 3.
3. Has it spent enough to judge?
If it has spent less than about one target cost per conversion, watch it. Don't block it yet.
This rule of thumb isn't a law. It just stops you judging a term on four clicks. Say your target is $80. A term that has had 30 clicks at $4 each has spent $120, which is more than one target with nothing to show for it. A term with 5 clicks at $4 has spent $20. That is a quarter of one conversion's budget, so the zero tells you almost nothing yet.
For low-converting, high-value accounts, such as B2B leads or big-ticket installs, raise the bar to two or three times the target. Conversions there are rare even on good terms.
If the term has spent enough, go to question 4.
4. Is the problem the term, or what happens after the click?
If the intent is right but the ad or page doesn't match it, keep it and fix the mismatch. If the intent is weak, block it.
Take "water heater replacement cost" in an emergency-repair ad group. The searcher may want a replacement. The ad talks about burst pipes and the page shows a phone number with no prices, so they leave. Blocking the term hides the symptom, but you sell water heater installs, so the term is not the problem. The routing is. Move it to the right ad group, or add it as a keyword where a better ad can serve it.
Compare "what size water heater do I need". It is close to the topic, but the person is researching, not hiring. If it has spent past your threshold with no conversions, block it. Question-style searches are where this call is hardest. The test before you block approach is worth reading if your account gets a lot of them.
5. Does it deserve its own keyword?
If a term has converted, repeats week after week and is currently matched by a broad or loosely related keyword, add it as a keyword. Use exact or phrase match, in the ad group whose ad and page fit it best.
This gives you control. You can set its bid on its own, write ad copy that repeats the searcher's words, and see its numbers without the noise from the broad keyword. Google generally prefers a keyword that is identical to the search term when choosing which keyword to serve. Even so, check next week's report to confirm the term now matches the new keyword.
If none of this applies, the term stays as it is: keep, with no action.
The borderline terms most checklists skip
These six kinds of term are where a weekly review turns into guesswork. Each one needs a fixed rule, written down once.
The term that converted once, at poor value
"cheap water heater install" converted once, for a $40 job, after $150 of spend. Question 1 says keep, and that is still right for negatives. One conversion means the intent can buy. The issue is value, and blocking is a blunt fix for a value problem. Put it on the watch list with a note to review it at the next bid or value check. If you track conversion value and use value-based bidding, the bidding can learn to pay less for it. That only works if the conversion value is recorded accurately.
Competitor names
"[competitor] plumbing reviews" or "[competitor] phone number". Most of these searchers are trying to reach a company they already chose, so they rarely convert. Some accounts do get customers this way, usually when the competitor is slow to answer the phone. Make one rule and apply it every time. Either block competitor names as a policy decision, or keep them and judge each one with the spend threshold like any other term. What wastes money is deciding case by case and changing your mind.
Your own brand in non-brand campaigns
Your brand name showing up in a generic Search campaign or in Performance Max is not a bad term. It is a term in the wrong place. Blocking it there is often right, but only if a brand campaign exists to catch it. If you block it without one, you can lose brand traffic. The Performance Max version of this problem has its own article: brand cannibalization in your search terms.
Partial matches with one wrong word
"tankless water heater repair" when you only service tank heaters. The term is 75% right. Block the wrong word, not the whole query. A phrase negative on "tankless" covers every future variant with one entry. Make sure the word is wrong in every context, though. "tankless" is safe if you never touch tankless units. "gas" is not safe if you service gas and electric. The phrase match negative examples show where these single-word blocks are safe and where they backfire.
Misspellings and close variants
"plumer near me" is the same intent as "plumber near me". Keep it. Misspellings feel like junk at a glance, which is why they get blocked by mistake. Judge them by the intent of the correctly spelled version.
High-spend terms that are too new to judge
Some terms appear suddenly and spend fast, like "burst pipe insurance claim plumber" after a cold snap. Last week they cost $90, and they were not in the account before. They have passed the spend threshold, but they appeared with an event and may fade with it. If the intent fits, keep the term on the watch list for one more week and don't block it in a hurry. A term that is still spending and still not converting a week later goes back through the tree with more data.
Choosing the match type and the level for each block
Use the narrowest negative that solves the problem, at the lowest level where the problem exists. A negative that is too wide quietly blocks traffic you wanted. You won't see that loss in any report.
Exact match blocks only that query and its close variants. Use it for single bad queries where the words themselves are fine in other contexts. "plumber open now cheap" might be junk as a whole, while "plumber open now" converts. Exact match is the safest block when you are unsure.
Phrase match blocks any query that contains the phrase in that order. Use it for a word or phrase that is wrong everywhere: "salary", "jobs", "tankless", "free course".
Broad match negatives block any query that contains all the words, in any order. They behave unpredictably across match types, so use them rarely and on purpose, never as a default.
For the level, pick the smallest one that fits:
- Ad group level when the term is fine for the account but wrong for that ad group. "water heater replacement" might be wrong for the emergency ad group and right for the installs ad group.
- Campaign level when the term is wrong for the whole campaign.
- Shared list when the term is wrong for the entire account: jobs, education, free. Shared lists work for Search and standard Shopping campaigns.
If a keyword suddenly loses impressions a week after a review, check your new negatives first. An over-wide block is the most common cause, and the fix is covered in keyword not showing impressions.
Where a weekly manual review falls short
A weekly checklist catches most of the waste in Search and Shopping. It has real gaps elsewhere, and you should know them before you rely on it.
Performance Max is harder. Search terms for Search, Shopping and Performance Max are all reported in campaign_search_term_view, so you can review them. Blocking them is another matter. A Performance Max campaign accepts up to 10,000 campaign-level negative keywords and does not accept shared negative keyword lists. Your account-wide "jobs, salary, free" list will not protect it. You have to add those negatives to each Performance Max campaign. If your negatives seem to have no effect there, this is usually why.
Some campaigns report no terms at all. Display, Video and App report no search terms, so this checklist cannot see their spend. You need placement reports and other tools for those.
Not every term is shown. The terms visible in the report often add up to less than the campaign's total cost. The gap is cost you cannot review by term, and no checklist can block what it cannot see.
Weekly is slow for fast-spending accounts. A bad term that starts on Monday has a full week to spend before you see it. In a small account that is a few dollars. In a large one it can be real money, and Smart Bidding learns from those clicks in the meantime. Junk queries skew the bids for the whole campaign, not just their own.
Blocking has a cost. Every negative is a bet that the term will never convert. The spend threshold and the 90-day conversion check lower the odds of a bad bet, but they don't remove them. That is why the watch list matters. It means you block on evidence rather than on a feeling.
What PPCleaner does with the same tree
PPCleaner automates the block side of this checklist. It does not handle the add-as-keyword side. Every night it reads the search terms your account paid for in Search, Shopping and Performance Max. It judges which ones were never going to convert and adds them as negative keywords.
It follows the same guardrail as question 1. It never blocks a term that converted in the last 90 days. In Performance Max it adds negatives at campaign level, because that is the only place they can go. Every block can be undone in one click for 90 days. If a borderline call goes the wrong way, reversing it takes one click and doesn't require a support ticket.
Here is what it won't do. It never touches budgets, bids, ads or keywords. So question 4 (fix the ad or page) and question 5 (add as keyword) stay with you. That leaves a shorter weekly review. You spend your 15 minutes on routing, new keywords and the pattern check, and the blocking runs every night instead of every Friday.
Frequently asked
How often should I review Google Ads search terms?
Weekly suits most accounts that spend a few thousand dollars a week. It is often enough to catch a bad term before it costs much, and there is enough data each week for the review to be worth doing. Accounts with high daily spend may need to look more often, or automate the blocking, because a bad term can spend for a full week before a weekly review sees it.
How much should a search term spend before I block it?
If the searcher could never buy (job seekers, DIY, free, wrong product), block it on the first click. If the intent could convert, a practical rule of thumb is to wait until the term has spent about one target cost per conversion with no conversions. For accounts with rare, high-value conversions, wait for two or three times that. This is a working rule, not a published Google figure.
Should I block a term that converted once but at a poor value?
Usually not with a negative keyword. One conversion shows the intent can buy, so the issue is value, not relevance. Put it on a watch list and handle it with bids or value-based bidding. Block it only if it keeps spending with no further conversions over a longer period.
Can I use the same checklist for Performance Max?
You can review Performance Max search terms the same way, because they are reported alongside Search and Shopping terms. Blocking works differently. Performance Max accepts up to 10,000 campaign-level negative keywords and does not accept shared negative keyword lists, so account-wide negatives have to be added to each Performance Max campaign separately.
When should a search term become a keyword instead of staying as it is?
When it has converted, shows up week after week, and is currently matched by a broad or loosely related keyword. Add it as an exact or phrase keyword in the ad group whose ad and landing page fit it best. You can then bid on it separately and write ad copy for it. Check the next report to confirm the term now matches the new keyword.
Clean your search terms. Save time and money.
Every night, PPCleaner reviews the searches your campaigns actually paid for, groups them by intent, and excludes the ones that never convert. No spend threshold to wait for, nothing touched but negatives — and one click undoes any of it.
Start 7 days freeSearch, Shopping and Performance Max · Cancel in one click
Related articles

Google Ads Reviews Searches Wasted? Measure Before Blocking
The usual advice is to add "reviews" and "complaints" as negatives. Some of these searches come from people one step from buying, so measure what they assist before you block them.

Misspelled Search Terms in Google Ads and Which to Block
Close variants send you typos of the queries you want and typos of the ones you don't. Most misspellings are harmless and convert fine. A few are a different search altogether, and your negative list won't catch them.

Google Ads Search Terms Report Filters That Surface Waste
Sorting the search terms report by cost puts your best terms at the top, not your waste. These six filter combinations, built around your target CPA, conversions and word count, show you what to block and what to leave alone.
