The working sequence.
12 sectionsFollow the document’s sequence, or jump to the step you need.
- 01Step 1: Search Term Cross-Reference (30 minutes)
- 02Step 2: Conversion Efficiency Comparison (15 minutes)
- 03Step 3: Brand vs. Non-Brand Segmentation (15 minutes)
- 04Step 4: Performance Max Channel Breakdown (15 minutes)
- 05Step 5: Brand Exclusion Checklist
View 2 more sections
Jump to a section
Put this guide to work
Cross-reference queries, brand/nonbrand evidence, channel mix and campaign roles, then investigate suspected overlap.
Your task: Map overlapping queries. Record the source, date range and reviewer beside the result. Leave anything you have not checked marked Not checked.
How to interpret it: A query appearing in multiple campaigns is an observation to investigate. It does not establish self-competition or lost revenue. Compare campaign roles, eligibility and the same observation window before proposing a change.
Related work: Plan feed and campaign segments with clear roles. · Design a test that can answer an incrementality question.
Before You Start
What you need:
- Google Ads account access with campaign-level reporting
- A spreadsheet application (Google Sheets or Excel)
- 90-120 minutes
What campaigns this applies to: This audit works for any account running 2 or more of the following campaign types simultaneously:
- Search campaigns
- Shopping campaigns (Standard)
- Performance Max campaigns
- Demand Gen campaigns
Multiple campaigns of the same type can also share query/product coverage. Scope the audit to the campaigns and reports actually available.
Step 1: Search Term Cross-Reference (30 minutes)
This is the core of the audit. You're looking for queries that appear in multiple campaign types.
Pull the Reports
For Search campaigns:
- Go to Google Ads > Campaigns > select your Search campaign(s)
- Click Keywords > Search Terms
- Set date range to last 30 days
- Export to CSV
For Shopping campaigns:
- Go to Google Ads > Campaigns > select your Shopping campaign(s)
- Click Keywords > Search Terms
- Set date range to last 30 days (same range as Search)
- Export to CSV
For Performance Max:
- Go to Google Ads > Campaigns > select your PMax campaign(s)
- Click Insights > Search Terms (note: PMax search term data is limited)
- Set date range to last 30 days
- Export to CSV
For Demand Gen (if applicable):
- Open its supported inventory, audience and conversion reports.
- Keep these observations separate from Search/Shopping/PMax query exports.
- Mark direct search-query coverage N/A; Demand Gen is not a keyword-search campaign.
Build the Cross-Reference
Open a new spreadsheet. Create these columns:
| Column A | Column B | Column C | Column D | Column E |
|---|---|---|---|---|
| Search Query | Reported in Search? | Reported in Shopping? | Reported in PMax? | Related Demand Gen research (not query coverage) |
Combine available query exports into Column A and deduplicate within each campaign type. Mark observed coverage only; missing source coverage stays Not checked. Column E records separate research or N/A, never invented Demand Gen query delivery.
Calculate Your Overlap Rate
Formula: Overlap Rate = (Queries appearing in 2+ campaign types) / (Total unique queries) x 100
**Your overlap rate: _______% **
How to read it: Record overlap as a descriptive rate within the exported report coverage. No universal overlap percentage establishes health, self-competition or financial loss; missing reports are unknown.
Step 2: Conversion Efficiency Comparison (15 minutes)
Compare observed outcomes on compatible dates, attribution and lag. Campaign traffic is selected differently; a CVR/CPA difference does not show the counterfactual result from transferring all traffic.
Build the Efficiency Table
For your top 50 overlapping queries (by impression volume), pull these metrics from each campaign type:
| Query | Search CVR | PMax CVR | Search CPA | PMax CPA | Search ROAS | PMax ROAS | Descriptive observation / limits |
|---|---|---|---|---|---|---|---|
How to compare: Record CVR, CPA and contribution under compatible windows. Prefer a descriptive comparison or an inconclusive result; selected traffic does not identify a counterfactual winner.
Count Your Winners
| Campaign Type | Comparable queries with higher observed metric | % of compared, mature query cases |
|---|---|---|
| Search | _______ | _______% |
| Shopping | _______ | _______% |
| PMax | _______ | _______% |
| Demand Gen | N/A: not keyword-query comparable | N/A |
This table identifies hypotheses for review. Do not infer optimal ownership from observed selected traffic or apply exclusions automatically.
Step 3: Brand vs. Non-Brand Segmentation (15 minutes)
Separate your overlapping queries into two groups: branded (queries containing your brand name) and non-branded.
Brand/nonbrand separation helps interpret existing demand. Reported credit is not incremental revenue; this workbook provides no account-specific incrementality factor.
Categorize Your Overlap
| Category | Count of Overlapping Queries | % of Total Overlap |
|---|---|---|
| Branded queries | _______ | _______% |
| Non-branded queries | _______ | _______% |
Measure Branded Overlap Cost
For branded overlapping queries specifically:
| Metric | Search | PMax |
|---|---|---|
| Branded query impressions | _______ | _______ |
| Branded query conversions | _______ | _______ |
| Branded query conversion rate | _______% | _______% |
| Branded query spend | $_______ | $_______ |
| Branded query CPA | $_______ | $_______ |
Branded review exposure: Record observed branded spend and outcomes, then formulate a controlled test. Spend on a lower-CVR campaign is not automatically waste or recoverable savings.
Step 4: Performance Max Channel Breakdown (15 minutes)
Use the available PMax channel report to inspect delivery and its coverage. A channel mix identifies questions; it does not establish cross-campaign competition or wasted spend.
Pull the Channel Data
In Google Ads, open your PMax campaign. Go to Insights and look for channel-level performance data. Note the percentage of spend and conversions by channel:
| Channel | % of PMax Spend | % of PMax Conversions | Overlap Risk |
|---|---|---|---|
| Search | _______% | _______% | Check actual shared query eligibility/report coverage |
| Shopping | _______% | _______% | Check shared product coverage and actual campaign roles |
| Display | _______% | _______% | Check actual shared audiences/inventory if applicable |
| YouTube | _______% | _______% | Check actual shared audiences/inventory if applicable |
| Discover | _______% | _______% | Check actual shared audiences/inventory if applicable |
| Gmail | _______% | _______% | Investigate the observed role; no fixed risk rating |
| Maps | _______% | _______% | Investigate actual eligibility and goals; no fixed risk rating |
Interpret the Results
Use channel mix to choose overlap questions, not to diagnose cannibalization from 30% or 20% cutoffs. Search/Shopping delivery can be an intentional PMax role; incrementality requires its own evidence.
Step 5: Brand Exclusion Checklist
Implementation Steps
-
1. Open PMax campaign settings. Go to your Performance Max campaign > Settings > Brand safety > Brand exclusions.
-
2. Review the selected brand entity/list. Apply it only where the agreed brand/non-brand role requires a supported exclusion.
-
3. Check observed variations. Verify how the supported brand control covers misspellings/abbreviations instead of assuming each is a separate keyword entry.
-
4. Review distinctive product names. Document which belong to the brand cohort and which supported control, if any, is appropriate.
-
5. Verify actual behavior. Allow for documented propagation/reporting lag and coverage limits; no fixed 24-hour check proves every brand query is excluded.
Brand Terms to Exclude
Record observed brand variations and coverage to review; this is not an instruction to enter every phrase into a brand-entity list:
| Type | Terms |
|---|---|
| Exact brand name | |
| Common misspellings | |
| Abbreviations | |
| Brand + "reviews" | |
| Brand + "discount" | |
| Brand + "coupon" | |
| Distinctive product names |
Post-Exclusion Monitoring (14 days)
Track these metrics for 14 days after implementing brand exclusions:
| Metric | Before Exclusion | Day 7 | Day 14 |
|---|---|---|---|
| PMax branded impressions | _______ | _______ | _______ |
| Search branded impression share | _______% | _______% | _______% |
| Branded conversion cost (Search) | $_______ | $_______ | $_______ |
| Total branded conversions (all campaigns) | _______ | _______ | _______ |
What to measure: Record actual brand coverage, cost and total business outcomes after lag. A before/after change without a suitable comparison does not prove prior credit-taking.
Step 6: Role Assignment Template
After the audit, assign each query (or query group) to the campaign type that should own it. This documents intended roles; actual coverage still needs readback.
The Four Campaign Roles
| Role | Campaign Type | Function | Budget Range |
|---|---|---|---|
| Intent Capture | Search | Own high-intent, non-branded queries | 15-30% of total |
| Product Discovery | Standard Shopping | Showcase products to comparing shoppers | 20-35% of total |
| Scale and Automation | PMax (brand excluded) | Discover new audiences and placements | 30-45% of total |
| Demand Generation | Demand Gen | Generate awareness in new audiences | 5-25% of total |
Query Assignment Worksheet
For the top 30 observed shared queries, document intended roles and any supported exclusion proposal. Apply only after relevance, coverage and outcome review; not every campaign type supports keyword exclusions.
| Query | Assigned To | Reason | Negative In |
|---|---|---|---|
Assignment decision logic:
Is the query supported by an available search-term report?
NO --> Record unknown coverage; do not invent a campaign assignment.
YES --> Is it branded?
YES --> Review intended brand roles and supported Search/PMax controls.
NO --> Review intent, product/page fit, economics and eligible inventory.
For each cohort, document a proposed role and how coverage will be checked.
Demand Gen research is separate; it is not a keyword-query routing destination.
Observed CVR/CPA and volume alone do not identify the best counterfactual owner.
Apply exclusions only after relevance, coverage, review and test-risk checks.
Negative Keyword Coordination
Coverage review: Relevant queries excluded for coordination may need another route; genuinely irrelevant queries need no positive coverage. If a relevant query is excluded from one route, verify whether the intended alternative is actually eligible to cover it. A positive keyword alone does not guarantee an impression, and a genuinely irrelevant query needs no replacement.
Use this tracker to verify coordination:
| Query Negated | Negated From | Actively Targeted In | Coverage Verified? |
|---|---|---|---|
| Y / N | |||
| Y / N | |||
| Y / N | |||
| Y / N | |||
| Y / N |
Step 7: Record Overlap Exposure and a Test Plan
| Query cohort | Observed spend and period | Coverage/lag limits | Hypothesis | Comparison and owner |
|---|---|---|---|---|
| Branded | ___ | ___ | ___ | ___ |
| Nonbranded | ___ | ___ | ___ | ___ |
Do not sum all lower-performing campaign spend as waste or multiply it by 12 as annual savings. Record actual controlled changes in cost, revenue and margin before estimating a commercial effect.
Action Plan
Based on your audit, prioritize fixes in this order:
This Week (Immediate)
- Review whether the agreed brand/non-brand roles require a supported exclusion and verify its actual scope (Step 5)
- Review demonstrated irrelevance and any routing proposal; selected-traffic CVR differences do not authorize negatives automatically
This Month (Short-term)
- Complete query-by-query role assignment for top 50 overlapping terms (Step 6)
- Build negative keyword coordination spreadsheet to prevent coverage gaps
- Review supported exclusion/list scope for any justified routing proposal; do not apply a list solely because a query overlaps
Ongoing (Monthly - 30 minutes)
- Re-run Step 1 (search term cross-reference) to catch new overlap
- Review brand exclusion effectiveness (branded impression share should stay high in Search)
- Check intended alternative eligibility for relevant excluded queries; irrelevant queries need no replacement
- Review actual brand-variant and emerging-query coverage under current negative/brand control behavior
Interpreting Common Patterns
After running the full audit, you'll likely see one of these patterns:
Pattern 1: High branded overlap (more than 50% of overlap is branded queries) Review the observed branded cohort. Next check: Review actual exclusion behavior and conversion lag; no fixed percentage or 14-day effect is established.
Pattern 2: PMax and Shopping competing on hero products Shared top-product coverage is a review flag, not proof of harmful competition. Custom labels can support product inclusion/exclusion groups, not an independent per-product priority bid. Review campaign-level targets, grouping, traffic and economics; PMax asset groups do not have independent bid strategies.
Pattern 3: Search and PMax competing on non-branded terms Observed shared non-brand queries need intent, eligibility, economics and coverage review. A 20% CVR difference on selected traffic does not identify a causal winner or justify reciprocal negatives; brief a bounded change and its comparison.
Pattern 4: Low overlap but poor overall performance Low observed overlap plus poor performance does not diagnose a missing funnel effect. Check report coverage, goals, product/page fit, economics and measurement before choosing an architecture change.
What This Workbook Doesn't Cover
This audit records overlap and briefs candidate changes; controlled evidence is needed to assess their effect. The complete orchestration system goes further:
- The 7-Layer Budget Distribution Model for allocating spend by funnel stage with margin-based guardrails
- The 4-Stage Deployment Model for adding campaign types sequentially without fragmenting data
- Incrementality testing protocols (brand exclusion tests, pause tests, geo holdout tests) for measuring whether campaigns generate real value or just claim credit
- The 5-Tier Negative Keyword Architecture with cross-campaign coordination and shared list management
- Presell page orchestration for integrating editorial and advertorial content into your paid media system
- Demand Gen coordination with applicable audience/measurement inputs; a fixed video-view window does not guarantee PMax routing or incremental demand
- The Monthly Orchestration Review protocol with specific trigger conditions for budget rebalancing
For the complete team-operated workflow, review the Google Ads AI Agentic System. The product page lists the current deliverables, price, license terms and setup responsibilities.
Choose who handles the next step.
Use the reference yourself, book a focused paid working session, or have Tegra handle the ongoing channel work.
Search, Shopping, Performance Max and Demand Gen campaigns, plus the research, product feeds, static and video creative, and landing, offer and pre-sale pages behind them.
Explore managed Google Ads → · See paid working sessions →