This page is updated every two months with current best practices for Google Ads negative keywords and search term management. As broad match and Smart Bidding take a bigger role, disciplined negative keyword management is one of the few direct levers you still have over where your budget is spent. Each update draws on our own experience plus authoritative industry sources and verified real-time research. Bookmark this page and check back for the latest negative keyword best practices. Each update includes worked examples with the arithmetic shown.
Last updated: 19 August 2026
In This Guide
- Executive Summary
- Benchmarks & Numbers at a Glance
- Why Negatives Matter with Broad Match
- Negative Match Types
- Account, Campaign & Ad Group Negatives
- Shared Negative Lists
- Working the Search Terms Report
- Brand, Competitor & Cross-Campaign
- Negatives & Smart Bidding
- Review Cadence
- Common Mistakes to Avoid
- What Changed Recently
- References
1. Executive Summary
Negative keyword management is one of the highest-leverage activities available to a Google Ads practitioner. The five principles below underpin every recommendation in this article.
- Negatives are a system, not a list. Effective management requires account-level, campaign-level, and ad group-level controls working together, supported by shared lists, rather than a single flat exclusion list applied everywhere.[1]
- Wasted spend is the primary diagnostic. Accounts with fewer than 1,000 negatives waste 28–42% of spend on irrelevant queries; top-quartile accounts waste only 5–12%.[57] Closing that gap is the measurable goal of this discipline.
- Broad match makes negatives more important, not optional. As broad match expands query coverage, negatives become the principal guardrail preventing irrelevant themes from consuming budget and corrupting Smart Bidding signals.[9][12]
- Review cadence determines outcomes. A weekly cost-sorted review of the Search Terms report, with daily checks during campaign launches, is the minimum standard for any active account.[1][3]
- Over-negation carries its own cost. Blocking too aggressively reduces conversion volume, starves Smart Bidding of learning signals, and can suppress legitimate traffic. Every exclusion decision requires deliberate intent.[8][12]
2. Benchmarks and Numbers at a Glance
| Metric | Typical range or threshold | Applies when | Source |
|---|---|---|---|
| Wasted spend — accounts with under 1,000 negatives | 28–42% of spend | B2B SaaS Google Ads accounts; 300+ accounts audited | [57] |
| Wasted spend — top-quartile accounts | 5–12% of spend | B2B SaaS Google Ads accounts; 300+ accounts audited | [57] |
| Industry-baseline negative keyword list size | 300–800 negatives | B2B SaaS accounts; represents median maturity | [57] |
| Top-quartile negative keyword list size | 1,500–8,000+ negatives | B2B SaaS accounts with active ongoing management | [57] |
| Wasted spend — unmanaged non-brand Search and Shopping | 15–30% of budget | Accounts with unmanaged or stale negative lists | [48] |
| Wasted spend — broad match without active monitoring | 30–50% of budget | Vendor claim; applies when broad match runs without a negative keyword review process | [60] |
| Wasted spend — average Google Ads account | 20–40% of budget | Vendor claim; applies across account types and verticals | [5] |
| Irrelevant impressions — top-performing accounts | <5% of impressions | Vendor claim; represents best-in-class negative management standard | [58] |
| Wasted spend before a structured negative keyword audit | ~35% of budget | Pre-audit state in a documented account review | [64] |
| Wasted spend after a structured negative keyword audit | ~8% of budget | Post-audit state in the same documented account review | [64] |
| Search Terms report review cadence — new campaign launches | Daily during early learning phase | Vendor claim; applies from launch until traffic pattern stabilises | [61] |
| Search Terms report review cadence — stable campaigns | At least weekly | Vendor claim; applies once campaign traffic is established | [61] |
| Negative keyword list audit cadence | Monthly | Vendor claim; applies to reviewing existing negatives for over-blocking | [61] |
Note: Where figures are marked as vendor claims, treat them as directional rather than statistically validated. The GrowthSpree figures are drawn from a study of 300+ B2B SaaS accounts and carry the most methodological weight in this table.[57]
3. Why Negative Keywords Matter More with Broad Match
Broad match in 2026 is the default match type Google recommends pairing with Smart Bidding. Its strength — reaching query variants a human keyword planner would not anticipate — is also its principal risk. Without a disciplined negative keyword framework, broad match will expand into adjacent intent categories, job-seeker queries, competitor product names, and how-to research traffic that has no commercial value for most advertisers.[9][10]
The relationship is asymmetric. Broad match magnifies the consequences of a poorly maintained negative list far more than exact or phrase match does, because exact match already constrains the query space to close variants of your chosen terms. Accounts running broad match without active monitoring waste 30–50% of budget on irrelevant traffic, according to vendor benchmark data.[60] By contrast, top-quartile accounts using well-structured negatives hold irrelevant spend to 5–12%.[57]
The practical implication is that broad match and negative keywords are not independent decisions. Choosing broad match without simultaneously investing in negative keyword infrastructure is a structural error. Negatives are the guardrails that allow broad match to discover genuinely useful query variants while preventing it from drifting into irrelevant intent.[12][14]
There is also a Smart Bidding dimension. Google’s automated bidding learns from the conversion signals it observes on the traffic you allow into the auction. If broad match surfaces high volumes of zero-converting, irrelevant queries before your negatives catch them, Smart Bidding accumulates misleading signals — effectively learning that certain query patterns are acceptable. Cleaning up the traffic pool with negatives is a prerequisite for reliable Smart Bidding performance, not an optional refinement.[6][12]
Worked example
Broad Match Waste Calculation for a Home Services Account
- Setup: A Melbourne home services account spending $10,000 per month runs all campaigns on broad match with a negative list of 220 terms added at account launch six months ago and not reviewed since.
- Numbers: At the lower bound of the broad-match-without-monitoring waste range (30%), the account wastes $3,000/month. At the upper bound (50%), waste is $5,000/month.[60] A top-quartile account at the same spend level with active negative management wastes 5–12% of $10,000, or $500–$1,200/month.[57] The gap between the worst-case scenario and best-practice performance is therefore $3,800–$4,500/month — $45,600–$54,000/year — on a $120,000 annual budget.
- Decision: Prioritise an immediate Search Terms report audit, move the account to a weekly review cadence, and target a negative list of at least 1,500 terms within 90 days.
- Why: The benchmark gap between under-managed broad match accounts (30–50% waste) and top-quartile accounts (5–12% waste) represents the financial case for investing in negative keyword infrastructure.[57][60]
4. Negative Match Types Explained
Google Ads supports three negative match types — broad, phrase, and exact — and each behaves differently from its positive counterpart. Understanding the documented behaviour of each is essential for predicting what a negative will and will not block.[15]
Negative Broad Match
A negative broad match keyword blocks your ad from showing when the search query contains all the terms in that negative keyword, regardless of word order.[7] It does not require the terms to appear together or in sequence. This makes it the most aggressive of the three types and the most likely to cause over-blocking if applied carelessly. Use it for themes that are universally irrelevant — for example, jobs, free, salary, tutorial — where any query containing these terms is almost certainly outside your target intent.[6][12]
Negative Phrase Match
A negative phrase match keyword blocks your ad when the search query contains the exact phrase, in the same word order, as the negative keyword. This gives you structural control over recurring unwanted query patterns. For most exclusions beyond universal junk categories, phrase match is the recommended default because it balances precision with coverage.[12]
Negative Exact Match
A negative exact match keyword blocks your ad only when the search query matches that specific term exactly (or is a very close variant). It carries the lowest risk of over-blocking but is also the easiest to under-apply, since query variants outside that exact form will not be blocked. Use it for one-off problem queries where you need surgical precision without disturbing adjacent traffic.[12]
One Behaviour Worth Noting
For search queries longer than 16 words, a negative keyword can still block the query if the negative term appears after the 16th word.[4] This matters when auditing long-tail queries and expecting that a negative should not have matched — it may have.
Match Type Comparison
| Negative match type | Blocking rule | Best use | Primary risk | Recommended default? |
|---|---|---|---|---|
| Broad | Blocks if all negative terms appear in the query, any order | Universal junk categories: jobs, free, DIY, salary | Over-blocking; can suppress valid queries containing the same words in a different context | No — use selectively |
| Phrase | Blocks if exact phrase appears in query, in order | Recurring unwanted query structures | Moderate; variants with different word order pass through | Yes — best general default |
| Exact | Blocks only the specific query (or very close variant) | One-off problem queries requiring surgical exclusion | Under-blocking; misses variants outside the exact form | Yes — for specific queries |
Worked example
Choosing the Right Match Type for a Software Account’s Job-Seeker Traffic
- Setup: A Brisbane B2B software account spending $8,000 per month is seeing 18% of its Search spend consumed by queries containing the word “jobs” or “careers” in various positions — for example, “project management software jobs”, “best project management software careers 2026”, and “project management jobs software engineer”.
- Numbers: 18% of $8,000 = $1,440/month on job-seeker traffic. At the industry baseline waste rate for under-managed accounts (28–42%), expected total waste is $2,240–$3,360/month.[57] Job-seeker waste alone accounts for $1,440 of that $2,240–$3,360 — approximately 43–64% of total waste.
- Decision: Add jobs and careers as negative broad match keywords to the account-level shared “Universal Exclusions” list. Do not use phrase or exact match here because any query containing these terms is irrelevant regardless of surrounding context.
- Why: Broad match negatives are appropriate when all queries containing the term are irrelevant; phrase or exact would leave variants unblocked and require ongoing patching of every new permutation.[6][12]
5. Account, Campaign and Ad Group Level Negatives
Structuring negatives across three levels is the architecture that separates a managed system from an ad hoc list. Each level serves a distinct purpose, and conflating them creates both over-blocking and maintenance complexity.[1][2][3]
Account-Level Negatives
Account-level negatives (applied via shared lists attached to all campaigns) are the foundation layer. They contain terms that are never relevant to any campaign in the account — universal junk traffic such as job, career, salary, free, DIY, tutorial, how-to, and unrelated industry terms. Centralising these exclusions here prevents repeated work at the campaign level and ensures that new campaigns are immediately protected.[1][2][7]
Campaign-Level Negatives
Campaign-level negatives handle intent shaping. These are terms that may be valid somewhere in the account but are wrong for a specific campaign’s objective. The most common use case is keeping brand traffic out of non-brand campaigns and generic terms out of brand campaigns. Campaign-level negatives are also used to prevent product-line cannibalism when two campaigns target adjacent audiences.[1][3][16]
Ad Group-Level Negatives
Ad group-level negatives should be used sparingly and only for surgical overlap control within tightly themed structures. Over-relying on this level creates maintenance burden and obscures the logic of the account over time. Reserve it for cases where two ad groups within the same campaign would otherwise compete for the same query.[3][9][13]
| Level | Scope | What belongs here | Management frequency |
|---|---|---|---|
| Account (shared list) | All campaigns | Jobs, salary, free, DIY, tutorial, unrelated industries, unserved locations | Monthly review; add immediately on discovery |
| Campaign | One campaign | Brand terms in non-brand campaigns; product lines in wrong campaigns; competitor terms where no conquest campaign exists | Weekly review during Search Terms audit |
| Ad group | One ad group | Overlap prevention between adjacent ad groups within the same campaign | As needed; audit quarterly to check continued necessity |
Worked example
Three-Level Structure for a Legal Services Account
- Setup: A Sydney legal services account spending $15,000 per month runs three campaigns: Brand, Personal Injury (PI), and Employment Law. The account has no formal negative structure — all negatives are mixed into campaign-level lists with no shared list.
- Numbers: Monthly spend: Brand $2,000, PI $8,000, Employment Law $5,000. At the unmanaged-list waste rate of 15–30%[48], total monthly waste is $2,250–$4,500. With a structured three-level approach targeting the top-quartile waste rate of 5–12%[57], waste would be $750–$1,800/month — a saving of $1,500–$2,700/month, or $18,000–$32,400/year.
- Decision: Create a shared “Universal Exclusions” list (containing: jobs, salary, free, DIY, law school, legal aid, template, form) applied to all three campaigns. Add brand terms as phrase match negatives at the campaign level in PI and Employment Law. Add “personal injury” and “injury claim” as phrase match negatives at the campaign level in Employment Law to prevent cross-campaign cannibalism.
- Why: Separating universal exclusions from intent-shaping exclusions means each layer can be maintained independently, reducing both waste and the risk of accidentally removing a strategically placed negative.[1][3]
6. Shared Negative Keyword Lists
Shared negative keyword lists are the most efficient mechanism for applying consistent exclusions across multiple campaigns without duplicating maintenance effort. In 2026, using shared lists for universal exclusions is standard practice among well-managed accounts.[1][4][17]
Google Ads allows you to create named shared lists in the Tools section and apply them to as many campaigns as needed. When you update a shared list, the change propagates to every campaign it is attached to simultaneously — a significant operational advantage over campaign-by-campaign management.[15]
Recommended Shared List Architecture
Rather than maintaining one monolithic shared list, organise exclusions into thematic lists. This makes auditing easier and allows you to attach specific lists only to the campaigns where they are relevant, rather than applying every exclusion globally.[1][4][19]
- Universal Junk List: Jobs, careers, salary, free, cheap, DIY, how-to, tutorial, YouTube, Reddit, Wikipedia. Attach to all campaigns.
- Brand Exclusions List: Your own brand terms in phrase and exact match. Attach to all non-brand campaigns.
- Competitor Exclusions List: Named competitor brands. Attach selectively — only to campaigns where you are not running a conquest strategy.
- Industry Irrelevants List: Terms specific to adjacent industries that are similar to your keywords but signal different intent. Attach to campaigns where that confusion risk exists.
- Geographic Exclusions List: Locations you do not service, if they appear in search queries. Attach to all campaigns or only to those affected.
One important limitation: if you manage a large account with many campaigns, verify current list caps and keyword limits within the Google Ads UI before building your architecture around any published number, since published third-party figures on platform limits can be outdated.[3]
Display and Video Campaigns
In Display and Video campaigns, Google treats all negative keywords as broad match regardless of how they are entered.[15] This makes shared list design especially consequential outside Search — a term added as phrase match on a shared list will function as broad match when that list is applied to a Display or Video campaign. Design Display-specific lists separately or document the match-type behaviour clearly for anyone managing those campaigns.
Worked example
Shared List Architecture for a Multi-Vertical E-Commerce Account
- Setup: A Melbourne e-commerce account selling tools and safety equipment spends $25,000 per month across 8 campaigns: 3 brand campaigns, 4 product campaigns (power tools, hand tools, safety footwear, workwear), and 1 Performance Max campaign. Each campaign currently has its own independently maintained negative list with significant duplication across lists.
- Numbers: The account has approximately 420 negatives spread across 8 campaign-level lists — well within the 300–800 baseline range but with heavy duplication.[57] A review finds that 280 of those 420 negatives (67%) are identical across 3 or more campaigns. Maintaining 280 duplicated negatives means any update requires changes in at least 3 places, creating 3× the error risk and 3× the maintenance time.
- Decision: Consolidate the 280 duplicated terms into a single “Universal Exclusions” shared list (including: free, cheap, hire, rental, DIY, tutorial, how-to, jobs, careers, apprenticeship, used, secondhand). Create a separate “Brand Terms” shared list and apply it to the 4 product campaigns and the PMax campaign. Retire duplicated campaign-level entries once shared lists are live and verified.
- Why: Shared lists eliminate duplication maintenance, ensure that a newly launched campaign inherits the full exclusion set immediately, and reduce the risk of inconsistent negative keyword application across the account.[1][4][17]
7. Working the Search Terms Report
The Search Terms report is the primary data source for identifying both negative keyword opportunities and new keyword additions. Working it effectively requires a defined process, not ad hoc browsing.[2][3][5]
Opening Protocol
Every review session should begin with the same sequence. Set the date range to the last 7 days for the working view — a 7-day lens catches new waste quickly without diluting the signal with stale data.[1][3] Sort by cost descending immediately. Sorting by impressions is a common mistake that draws attention to high-volume queries that may cost little, while expensive irrelevant queries with lower impression counts remain invisible.[1][6]
What to Look For
- Zero-conversion spend above target CPA: Any query that has spent more than your target CPA with zero conversions is a strong negative candidate. Flag it immediately.[1][6][15]
- Off-topic intent: Queries that contain your keywords but signal research, job-seeking, competitor comparison, or unrelated-category intent. These require manual judgement — the report cannot determine relevance automatically.[1][3][14]
- Recurring waste patterns: The same theme appearing across many different queries — for example, multiple variants of “how to”, “tutorial”, or a competitor name — is a signal that a shared list needs updating rather than individual negatives being added one by one.[3][8]
- Converting queries not yet targeted: Search terms generating conversions that do not match any existing keyword exactly should be evaluated for promotion to exact match keywords with dedicated ad copy.[1][3][15]
N-Gram Analysis for Scale
For accounts with high query volume, manual row-by-row review is insufficient. N-gram analysis — examining which individual words or two-word combinations appear most frequently across costly, non-converting queries — allows pattern-based negative identification at scale. A query might individually be borderline, but if the word “training” appears in 40 zero-converting queries across the last 30 days, that is a systemic waste pattern, not a one-off.[4]
Promoting Converting Queries to Keywords
When a search term generates conversions and is not already captured by an exact match keyword, evaluate whether it deserves dedicated control. Promote it to exact match if it has sufficient volume and commercial intent to justify its own keyword, ad copy, and landing page alignment.[1][15] Do not promote every converting query automatically — assess whether dedicated control is worth the structural complexity.
Worked example
Weekly Search Terms Review for a Dental Practice Account
- Setup: A Sydney dental practice account spending $4,500 per month on Google Ads runs 2 campaigns (general dentistry and cosmetic dentistry) on phrase and broad match. The account manager conducts a weekly Search Terms review every Monday using a 7-day date range sorted by cost descending.
- Numbers: In the review for 7–13 October 2026, the report shows 94 unique search terms. The top 10 by cost account for $1,260 (28% of weekly spend of $4,500 ÷ 4.3 weeks = $1,047/week; actual weekly spend this period: $1,260 across top 10 terms). Of those 10, 3 terms have zero conversions and a combined cost of $380: “dental school near me” ($180), “dentist jobs Sydney” ($140), “free dental check-up” ($60). Target CPA is $120. Each zero-conversion term has spent above $120 with no conversion.
- Decision: Add “dental school” as a phrase match negative at the campaign level (general dentistry). Add “dentist jobs” as a phrase match negative to the Universal Exclusions shared list (applies to both campaigns). Add “free dental” as a phrase match negative to the Universal Exclusions shared list. Document each addition with the date (13 October 2026), the cost-to-date figure, and the reason in the account’s negative keyword log.
- Why: Any query spending above the $120 target CPA with zero conversions meets the threshold for immediate negation; sorting by cost first ensures the highest-waste queries are actioned before lower-cost borderline cases.[1][6][15]
Worked example
N-Gram Analysis to Find a Systemic Waste Pattern
- Setup: A Perth HR software account spending $12,000 per month runs broad match across 5 ad groups. A 30-day Search Terms export (1 September–30 September 2026) produces 640 unique search terms. Manual row-by-row review is impractical at this volume.
- Numbers: N-gram analysis of the 640 terms identifies that the word “template” appears in 47 unique queries (7.3% of all queries) with a combined cost of $890 and zero conversions. The word “course” appears in 31 queries with a combined cost of $640 and zero conversions. The word “certification” appears in 22 queries with a combined cost of $510 and zero conversions. Total waste from these three patterns: $2,040 — 17% of the $12,000 monthly budget.
- Decision: Add “template”, “course”, and “certification” as negative broad match keywords to the Universal Exclusions shared list, effective 1 October 2026.
- Why: When a single word appears across 22+ zero-conversion queries with combined spend exceeding the account’s target CPA, it is a systemic pattern requiring a shared-list exclusion rather than individual query-by-query negation.[4][57]
8. Brand, Competitor and Cross-Campaign Negatives
Brand and competitor negative keyword decisions are strategic, not mechanical. Unlike universal junk exclusions, these require deliberate choices about campaign intent, budget allocation, and commercial objectives before any negative is applied.[1][3][18]
Brand vs Non-Brand Separation
The most common and most important application of campaign-level negatives is preventing brand traffic from entering non-brand campaigns. When a user searches your brand name and your broad match non-brand campaign captures that query, you pay non-brand CPCs for a query you should be winning cheaply on brand terms, and you corrupt your non-brand performance data with high-converting brand traffic. The fix is to add your brand terms (and common misspellings) as phrase match negatives in all non-brand campaigns.[3][16][18]
The reverse also applies: generic category terms should typically be excluded from brand campaigns to keep brand campaigns focused on navigational and brand-aware queries, and to prevent them from cannibalising non-brand campaign data.[1][18]
Competitor Negatives
Competitor term management is a strategic decision, not a default setting. The two scenarios are mutually exclusive:
- No conquest strategy: Add competitor brand names to a Competitor Exclusions shared list and apply it to all campaigns. This prevents spend on traffic with no plausible conversion pathway and keeps your impression share data clean.[1][4][6]
- Active conquest strategy: Do not apply competitor names as account-wide negatives. Instead, isolate them in a dedicated conquest campaign where you can manage bids, messaging, and budget separately. Apply the competitor exclusion list only to your non-conquest campaigns to prevent leakage.[4][18]
Do not make competitor terms account-wide negatives if you run dedicated conquest campaigns — doing so would block your own conquest strategy.[4][18]
Cross-Campaign Cannibalism
In multi-campaign accounts, broad match can cause one campaign to capture traffic intended for another. The diagnostic is straightforward: if the same search query appears in the Search Terms report for two different campaigns in the same week, one of them is being served incorrectly. Use campaign-level negatives to route each query to its intended campaign — typically by adding the overlapping term as a phrase match negative in the lower-priority campaign.[1][3][9]
Worked example
Brand Negative to Prevent Non-Brand Campaign Contamination
- Setup: A national accounting software company spending $20,000 per month runs a Brand campaign ($3,000/month) and a Non-Brand Acquisition campaign ($17,000/month). A weekly Search Terms review for 1–7 November 2026 shows that the Non-Brand campaign captured 38 queries containing the brand name “ClearLedger” (anonymised) at a total cost of $740 and a conversion rate of 14.5% — significantly above the non-brand campaign average of 3.2%.
- Numbers: The 38 brand queries in the Non-Brand campaign generated 5.5 attributed conversions ($740 ÷ $134 average CPC = 5.5 clicks; 5.5 × 14.5% = 0.8 conversions — rounding to 1 conversion for illustration). However, these queries inflate the Non-Brand campaign’s CPA, making it appear to perform better than it does on genuine non-brand traffic. Removing brand traffic from the Non-Brand campaign would reduce its apparent conversion rate from 3.2% to approximately 2.8% and expose the true non-brand CPA: $17,000 ÷ (17,000/134 × 2.8%) = approximately $281 true non-brand CPA vs $254 blended CPA.
- Decision: Add “ClearLedger” and its two common misspellings (“ClearLedge”, “ClearLeger”) as phrase match negatives at the campaign level in the Non-Brand Acquisition campaign, effective 8 November 2026.
- Why: Brand queries in non-brand campaigns inflate conversion rate and suppress true CPA, making Smart Bidding optimise against a misleading signal; separating brand and non-brand traffic is the prerequisite for accurate performance measurement.[3][16][18]
9. Negatives and Smart Bidding
Smart Bidding and negative keyword management have a reciprocal relationship that is frequently misunderstood. The common misconception is that Smart Bidding reduces the need for negatives because the algorithm will “learn” to avoid irrelevant queries. This is incorrect. Smart Bidding optimises bid decisions for queries that enter the auction — it does not prevent irrelevant queries from entering the auction in the first place. That function belongs entirely to negative keywords.[9][12][14]
How Dirty Traffic Harms Smart Bidding
When irrelevant queries enter the auction and receive clicks, Smart Bidding records those interactions as part of its learning data. If irrelevant queries generate zero conversions, the algorithm eventually learns to bid lower on those patterns — but this takes time and wastes budget during the learning period. More problematically, if irrelevant queries occasionally generate an accidental conversion (for example, a job-seeker who also happens to purchase), the algorithm may incorrectly reinforce those query patterns as valuable. Clean traffic pools are a prerequisite for reliable Smart Bidding, not a refinement applied after the bidding is working well.[6][8][12]
The Over-Negation Risk
The counterbalancing risk is real: adding too many negatives reduces the total conversion volume available for Smart Bidding to learn from. Most Smart Bidding strategies require a minimum of 30–50 conversions per month per campaign to move out of the learning phase and into stable optimisation. If aggressive negating reduces your conversion volume below that threshold, Smart Bidding will remain perpetually in learning mode, producing volatile CPAs and unreliable delivery.[8][12]
The practical implication is to prioritise negatives that address clear intent mismatches (zero-converting queries above target CPA) rather than blocking borderline queries that occasionally convert. When in doubt, monitor before blocking.[6][12]
Negatives in Performance Max
Performance Max campaigns in 2026 support account-level negative keyword lists. Applying your Universal Exclusions shared list to PMax is a minimum baseline — PMax’s broader inventory reach across Search, Display, YouTube, Gmail, and Maps makes exclusion management even more consequential than in standard Search campaigns.[19] Campaign-level negative keywords within PMax are available through Google Ads support request for established accounts; verify current UI availability within your account before relying on published information about PMax negative access.
Worked example
Over-Negation Reducing Smart Bidding Below Learning Threshold
- Setup: An Adelaide B2B SaaS account spending $9,000 per month on a single lead generation campaign runs Target CPA bidding at $180. The account manager, following an aggressive negative keyword audit in September 2026, adds 340 new negatives in a single week, including several broad match exclusions for terms like “software”, “platform”, and “system” added in an attempt to block research-intent traffic.
- Numbers: Before the audit, the campaign generated 48 conversions per month (rate: $9,000 ÷ $180 = 50 conversions; actual 48). After the bulk negation, impression volume drops 44% and conversion volume falls to 19 conversions in the first 30 days post-change — 19 conversions is 38% of the 50 needed to sustain stable Target CPA learning. The campaign enters the learning phase indicator in the Google Ads UI and CPA rises to $473 ($9,000 ÷ 19) against a $180 target.
- Decision: Remove the 3 broad match negatives (“software”, “platform”, “system”) immediately. Retain the remaining 337 specific phrase and exact match negatives. Set a monitoring rule: if monthly conversions fall below 35 (70% of the 50-conversion threshold), pause any pending negative additions until volume recovers.
- Why: Smart Bidding requires sufficient conversion volume to exit the learning phase; broad match negatives on high-frequency category terms can suppress impression volume to the point where the algorithm cannot function reliably.[8][12]
10. Review Cadence and Workflow
A defined review cadence transforms negative keyword management from a reactive task into a systematic discipline. The right cadence depends on account spend level, campaign age, and match type mix.[1][3][5]
Recommended Cadence by Account Stage
| Account stage | Search Terms report cadence | Negative list audit cadence | Date range to use |
|---|---|---|---|
| New campaign launch (first 2 weeks) | Daily | Daily review of new additions | Last 24–48 hours |
| Active campaign, broad or phrase match | Weekly | Monthly | Last 7 days for working view |
| High-spend or high-velocity (>$20,000/month) | Twice-weekly | Monthly | Last 7 days |
| Stable, lower-spend account (<$3,000/month) | Weekly minimum; bi-weekly acceptable | Quarterly | Last 14–30 days |
| New campaign launch (weeks 3–8) | Weekly | Monthly | Last 7 days |
Workflow for Each Review Session
- Step 1 — Open Search Terms report, set to last 7 days, sort by cost descending. This immediately surfaces where budget went, not where impressions went.[1][6]
- Step 2 — Flag zero-conversion queries above target CPA. These are your highest-priority negatives. Action them before anything else.[1][6][15]
- Step 3 — Scan for intent mismatches. Read the top 50 queries by cost for obvious off-topic intent that the algorithm cannot detect automatically.[1][3]
- Step 4 — Identify recurring patterns. If the same root word or phrase appears in 5 or more non-converting queries, consider a shared list addition rather than individual negatives.[3][8]
- Step 5 — Review converting queries for keyword promotion. Flag search terms generating conversions that are not matched by an existing exact match keyword for potential promotion.[1][3][15]
- Step 6 — Log every change. Record the date, the term added, the match type, the level (account/campaign/ad group/shared list), and the reason. This log is essential for later audits and for avoiding the removal of strategically placed negatives.[1][3]
Monthly Negative List Audit
Monthly audits of existing negative lists serve a different purpose from weekly Search Terms reviews. The monthly audit checks for over-blocking: terms that were added as negatives months ago and may now be suppressing valid traffic as your product range, target audience, or market conditions change. What was irrelevant last quarter may be a legitimate query today.[6][8][61] Use Change History to correlate performance dips with negative keyword additions and investigate any case where a significant impression or conversion drop followed a bulk negative addition.[4][5]
Worked example
Monthly Negative Audit Identifying an Over-Blocking Error
- Setup: A Canberra cybersecurity consultancy account spending $6,000 per month added “training” as a negative broad match keyword to its Universal Exclusions shared list in March 2026 to block job-seeker and course-seeker traffic. In a monthly audit conducted on 1 August 2026, the account manager reviews Change History and notices that impressions for the “security awareness” ad group dropped 31% in the week of 10 March 2026, coinciding exactly with the “training” broad match negative going live.
- Numbers: Pre-March impressions for “security awareness” ad group: approximately 1,400/week. Post-March impressions: approximately 970/week — a drop of 430 impressions/week, or 1,720/month. At a historical CTR of 4.2% and CPC of $18.50, those lost impressions represent approximately 72 lost clicks/month and $1,332 in suppressed spend. The account’s target CPA is $280, and the pre-March conversion rate was 6.8%, meaning the suppressed clicks would have generated approximately 4.9 conversions/month — worth $1,372 at target CPA.
- Decision: Remove “training” from the Universal Exclusions shared list. Replace it with two phrase match negatives at the campaign level: “security training course” and “cybersecurity training certification”, which specifically target course-seeker queries without blocking the commercially relevant “security awareness training for employees” query type.
- Why: Monthly audits using Change History are the mechanism for detecting over-blocking caused by broad match negatives; the fix is to replace broad negatives with more precise phrase negatives that target the specific unwanted pattern.[6][8][61]
11. Common Mistakes to Avoid
The following errors appear consistently across account audits in 2026. Each one is avoidable with a defined process and deliberate decision-making.[1][3][9][12]
Adding Negatives Without Logging the Reason
In multi-manager accounts, negatives added without documentation create compounding problems. A term blocked for a specific strategic reason six months ago may be removed by a new team member who sees it as incorrect. The log does not need to be complex — date, term, match type, level, and a one-sentence reason is sufficient, but it must be maintained consistently.[1][3]
Using Broad Match Negatives Indiscriminately
Broad match negatives applied to common words can suppress large volumes of legitimate traffic without any visible alert in the UI. A single broad match negative for a word like “system”, “service”, or “solution” in a B2B account can block hundreds of relevant queries silently. Reserve broad match negatives for terms that are irrelevant in every possible context.[7][12]
Reviewing by Impressions Instead of Cost
Sorting the Search Terms report by impressions directs attention to high-frequency queries that may be cheap and relatively harmless, while expensive irrelevant queries with lower impression counts remain unaddressed. Always sort by cost first.[1][6][15]
Neglecting New Campaigns in the First 72 Hours
New campaigns on broad or phrase match can generate significant irrelevant traffic within 48–72 hours of launch. Waiting a week to conduct the first Search Terms review means budget has already been wasted on queries that a prompt audit would have caught. New campaigns require daily review during the first two weeks.[5][61]
Applying Competitor Negatives Account-Wide When Running Conquest Campaigns
Adding a competitor’s brand name to an account-level negative list blocks every campaign, including any dedicated conquest campaign targeting that competitor’s audience. Competitor negatives must be applied at the campaign level with deliberate attention to which campaigns they are and are not attached to.[4][18]
Never Revisiting Old Negatives
A negative keyword added in January may be suppressing a query pattern that became commercially relevant by July as your product range, pricing, or target market evolved. Treating the negative list as permanent once established is one of the most common causes of unexplained impression decline in mature accounts.[6][8][61]
Over-Negating to the Point of Starving Smart Bidding
Aggressive bulk negation that pushes monthly conversion volume below 30–50 conversions per campaign will trigger Smart Bidding learning instability, producing volatile CPAs and erratic delivery. Every negative added during an active Smart Bidding campaign should be evaluated for its likely impact on total conversion volume, not just its impact on individual query relevance.[8][12]
Worked example
The Cost of Not Reviewing a New Campaign for 7 Days
- Setup: A Gold Coast real estate agency launches a new Google Ads campaign on 3 November 2026, spending $200/day, targeting “property management” keywords on broad match. The account manager schedules the first Search Terms review for 10 November 2026 — 7 days after launch.
- Numbers: Total spend over the 7 days before the first review: $1,400. A post-review audit on 10 November finds that 34% of that spend — $476 — went to queries including “property management jobs”, “property manager salary Gold Coast”, “property management courses”, and “free property management software”. All 34% of spend is zero-conversion. Had a daily review been conducted from day 1 and these patterns negated by day 2, the estimated avoidable waste (days 2–7, 6 days × $200 × 34%) = $408. The $476 is not entirely recoverable, but $408 of it was preventable with a day-2 review.
- Decision: For all future campaign launches, schedule a Search Terms review within 48 hours of launch (by 5 November for any campaign launching 3 November). Add “property management jobs”, “property manager salary”, “property management courses”, and “free property management” as phrase match negatives to the campaign-level list on day 2.
- Why: New broad match campaigns can generate irrelevant query traffic within 24–48 hours of launch; the vendor benchmark cadence recommends daily review during the early learning phase precisely to prevent this avoidable waste.[5][61]
12. What Changed Recently (Last 30 Days)
As of August 2026, there is no single documented Google Ads product launch or formal policy change specifically targeting negative keywords or the Search Terms report in the last 30 days. What the available sources confirm is a set of reinforced operational practices and clarified platform behaviours that practitioners should be aware of.[4][5]
Reinforced Platform Behaviours
- Search Terms report remains the primary source for negative keyword discovery. Google’s help documentation continues to recommend using the report to identify irrelevant queries and adding them as negatives directly from the report interface.[14][15]
- Change History is the built-in audit mechanism. When investigating a performance shift, use Change History to correlate the shift with newly added negative keywords within the same date range. This remains the only native tool within Google Ads for this purpose — there is no dedicated “negative keywords added” report available through the standard UI or API in the materials reviewed.[4][5]
- Negative broad match behaviour is unchanged. For negative broad match, your ad will not show if the search contains all negative keyword terms, even if they appear in a different order. This documented behaviour has not changed but remains the most commonly misunderstood aspect of negative match type mechanics.[7]
- Negative match types remain editable in the UI. You can edit the match type of an existing negative keyword within the Google Ads interface without deleting and re-adding the term.[6][15]
- The 16-word query rule applies. For search queries longer than 16 words, a negative keyword can still match and block the query if the negative term appears after the 16th word. This behaviour is documented but easy to overlook when auditing unexpected blocks on long-tail queries.[4]
Operational Emphasis Shifts in the Last 30 Days
The most recent practitioner guidance (published in July–August 2026) emphasises three operational adjustments over previous recommendations:[57][61][64]
- Review the last 30 days of search terms, not just the last 7 days, when conducting monthly audits. A 7-day lens is appropriate for weekly working reviews, but the monthly audit should use a 30-day window to catch slower-developing waste patterns and seasonal shifts.[4]
- Prioritise over-blocking audits, especially for accounts that were aggressively negated in Q1–Q2 2026. Several accounts that underwent bulk negative audits earlier in 2026 are now showing signs of suppressed impression share in categories that have since become commercially relevant. Monthly audits using Change History are the diagnostic tool.[7][9]
- Maintain a running log outside the Google Ads UI. Google Ads does not provide a simple exportable report of “all negative keywords added in a date range” from Change History. External documentation — a shared spreadsheet or project management log — remains the only reliable way to maintain institutional knowledge about why specific negatives were added.[1][4]
Worked example
Using Change History to Diagnose a Post-August 2026 Impression Drop
- Setup: A Hobart accountancy firm account spending $3,500 per month notices that impressions in its “tax return” campaign dropped from approximately 2,200/week to 1,350/week between 4 August and 11 August 2026 — a 38.6% drop with no change in bids, budget, or Quality Score during the period.
- Numbers: Impression drop: 2,200 − 1,350 = 850 impressions/week. At a historical CTR of 5.1%, that represents approximately 43 lost clicks/week. At an average CPC of $4.20, the lost clicks represent $181/week in suppressed spend — $724/month. At a historical conversion rate of 7.3%, the suppressed clicks represent approximately 3.1 fewer conversions/week, or 12.4/month. At a target CPA of $85, that is $1,054/month in missed conversion opportunity.
- Decision: Open Change History, filter to 1–11 August 2026, and review all negative keyword changes. A review identifies that “accounting” was added as a negative broad match keyword on 5 August 2026 — intended to block “accounting jobs” queries — but its broad nature is blocking “tax return accounting services” and “online accounting tax return” as well. Remove “accounting” from the broad match negatives and replace with “accounting jobs” (phrase match) and “accounting courses” (phrase match) as campaign-level negatives, effective 12 August 2026.
- Why: Change History is the only native mechanism for correlating a performance shift with a specific negative keyword change; broad match negatives on category-level terms silently suppress valid traffic without any UI alert.[4][5][7]
References
- [1] https://www.keywordme.io/blog/negative-keyword-list-google-ads-best-practices www.keywordme.io
- [2] https://verdemedia.com/blog/google-ads-keyword-strategy-2026 verdemedia.com
- [3] https://adssensor.com/blog/negative-keyword-strategy adssensor.com
- [4] https://www.capconvert.com/learn/blog/negative-keyword-strategy-for-2026-lists-scripts-and… www.capconvert.com
- [5] https://www.groas.com/post/negative-keywords-for-google-ads-the-complete-2026-list-500-key… www.groas.com
- [6] https://searchengineland.com/negative-keywords-strategy-476563 searchengineland.com
- [7] https://www.groas.com/post/google-ads-negative-keyword-strategy-2026-campaign-level-ai-max… www.groas.com
- [8] https://directiveconsulting.com/blog/the-b2b-marketers-guide-to-google-ads-best-practices-… directiveconsulting.com
- [9] https://www.karooya.com/blog/negative-keywords-in-google-ads-2026-are-you-using-them-to-fi… www.karooya.com
- [10] https://searchengineland.com/google-ads-mistakes-avoid-449288 searchengineland.com
- [11] https://www.get-ryze.ai/blog/google-ads-keyword www.get-ryze.ai
- [12] https://www.optmyzr.com/blog/negative-keywords/ www.optmyzr.com
- [13] https://www.keywordme.io/blog/negative-keywords-google-ads-2098f www.keywordme.io
- [14] https://support.google.com/google-ads/answer/2453972?hl=en support.google.com
- [15] https://support.google.com/google-ads/answer/2453983?hl=en support.google.com
- [16] https://gofishdigital.com/blog/negative-keywords-in-google-ads-the-key-to-scalable-campaig… gofishdigital.com
- [17] https://twominutereports.com/blog/google-ads-best-practices twominutereports.com
- [18] https://support.google.com/google-ads/thread/413560562/google-ads-best-practices-for-2026?… support.google.com
- [19] https://blog.adnabu.com/google-ads/performance-max-negative-keywords/ blog.adnabu.com
- [20] https://www.reddit.com/r/googleads/comments/1pdhpd8/what_are_your_google_ad_tips_for_2026/ www.reddit.com
- [21] https://darlingtongrowth.com/guides-google-ads-search-terms-report darlingtongrowth.com
- [22] https://www.keywordme.io/blog/google-ads-search-term-report-optimization www.keywordme.io
- [23] https://stanconsultingllc.com/learn/google-ads-diy/how-to-assess-search-terms-review stanconsultingllc.com
- [24] https://advantrise.com/search-terms-report-in-google-ads-the-audit-that-saves-your-budget advantrise.com
- [25] https://buttercup.digital/resources/google-ads-search-terms-report buttercup.digital
- [26] https://www.keywordme.io/blog/ppc-search-term-management-best-practices www.keywordme.io
- [27] https://lionelz.com/en/blog/search-terms-google-ads/ lionelz.com
- [28] https://almcorp.com/blog/how-voice-search-is-changing-google-ads-search-term-reports/ almcorp.com
- [29] https://searchengineland.com/google-ads-search-terms-report-tips-465174 searchengineland.com
- [30] https://support.google.com/google-ads/answer/2472708?hl=en support.google.com
- [31] https://www.keywordme.io/blog/what-is-search-term-report-optimization www.keywordme.io
- [32] https://www.databloo.com/blog/best-google-ads-reports/ www.databloo.com
- [33] https://support.google.com/google-ads/answer/2684537?hl=en support.google.com
- [34] https://www.northcountrygrowth.com/blog/google-ads-search-terms-report-how-to-actually-use… www.northcountrygrowth.com
- [35] https://blog.adnabu.com/google-ads/google-ads-search-term-report/ blog.adnabu.com
- [36] https://dashthis.com/blog/google-ads-best-practices/ dashthis.com
- [37] https://www.keywordme.io/blog/google-ads-search-terms-best-practices www.keywordme.io
- [38] https://support.google.com/google-ads/thread/413560562/google-ads-best-practices-for-2026?… support.google.com
- [39] https://support.google.com/google-ads/answer/16668865?hl=en support.google.com
- [40] https://support.google.com/google-ads/answer/7102995?hl=en support.google.com
- [41] https://groups.google.com/g/adwords-api/c/GeAwrD4rayc/m/5Pq7F44TDAAJ groups.google.com
- [42] https://groups.google.com/g/adwords-api/c/403HGadkQwc groups.google.com
- [43] https://support.google.com/google-ads/answer/9701952 support.google.com
- [44] https://groups.google.com/g/adwords-api/c/GeAwrD4rayc/m/lVxGHZ_-DAAJ groups.google.com
- [45] https://support.google.com/google-ads/answer/19888?hl=en support.google.com
- [46] https://support.google.com/google-ads/answer/7102298?hl=en support.google.com
- [47] https://www.keywordme.io/blog/how-to-audit-negative-keyword-performance www.keywordme.io
- [48] https://opascope.com/insights/negative-keywords/ opascope.com
- [49] https://support.google.com/google-ads/answer/7102466?hl=en support.google.com
- [50] https://www.reddit.com/r/PPC/comments/ce78yx/is_there_any_way_to_check_the_change_history_… www.reddit.com
- [51] https://www.digitalposition.com/resources/blog/seo/googles-negative-keyword-list-limit-may… www.digitalposition.com
- [52] https://developers.google.com/search-ads/v2/report-types/negativeCampaignKeyword?authuser=… developers.google.com
- [53] https://www.keywordme.io/blog/how-to-use-negative-keywords-in-google-ads www.keywordme.io
- [54] https://support.google.com/google-ads/editor/answer/30553?hl=en support.google.com
- [55] https://www.reddit.com/r/PPC/comments/vo4ykk/how_frequently_should_i_check_for_negative/ www.reddit.com
- [56] https://growleads.io/blog/negative-keywords-google-ads/ growleads.io
- [57] https://www.growthspreeofficial.com/blogs/b2b-saas-negative-keyword-list-size-benchmarks-2… www.growthspreeofficial.com
- [58] https://www.keywordme.io/blog/google-ads-irrelevant-search-terms-eating-budget www.keywordme.io
- [59] https://ad-times.com/how-to-cut-wasted-spend-in-google-ads-a-complete-2026-guide/ ad-times.com
- [60] https://www.optimyzee.com/blog/google-ads-wasted-budget-search-terms-ai www.optimyzee.com
- [61] https://www.keywordme.io/blog/how-often-should-you-review-search-terms-report www.keywordme.io
- [62] https://www.growthspreeofficial.com/blogs/b2b-saas-google-ads-search-term-relevance-benchm… www.growthspreeofficial.com
- [63] https://www.squared.io/case-studies/suzuki-the7stars www.squared.io
- [64] https://www.arjankc.com.np/blog/google-ads-cpa-reduction-case-study-nepal-2026/ www.arjankc.com.np
- [65] https://ad-times.com/how-to-fix-wasted-spend-in-google-ads-a-complete-2026-guide/ ad-times.com
- [66] https://clicksgeek.com/wasted-ad-spend-on-google-ads/ clicksgeek.com
- [67] https://www.meticulosity.com/blog/identify-seasonal-trends-with-adwords-search-terms-repor… www.meticulosity.com
- [68] https://adsbot.co/negative-keywords-in-google-ads/ adsbot.co
- [69] https://ppcchief.com/resources/negative-keyword-list ppcchief.com
Don’t want to do this yourself?
Get a complete, ready-to-launch Google Ads campaign — keywords validated against real search data, copy written to the best practices on this page, and a proper negative-keyword list — built for your business and delivered within 24 hours.
One-off price · Human-reviewed · No subscription · No access to your Google Ads account
This page is maintained by Sean Cooney at Omologist.com. Content is refreshed every two months using real-time research from authoritative Google Ads sources. Worked examples are illustrative scenarios calculated from published benchmarks, not client results.

