Performance Max overlap is not automatically cannibalization. Google says Performance Max can overlap with other campaigns when goals, bidding targets, and settings make the same inventory eligible; an exact-match Search keyword can be preferred in the relevant auction. The correct question is whether the interaction is incremental, understood, serviceable, and measured against the right outcome.
Google’s Performance Max search terms report exposes search terms and associated landing pages, while its results guidance recommends reviewing channel performance, conversion lag, search impact, and changes. These sources support an audit method, not a promise that exclusions or a specific campaign type will improve results.
State the decision
Decide what the audit must answer: whether to protect branded Search, diagnose budget substitution, improve query control, change goals, add exclusions, split a campaign, or pause a route. Define account, market, service, period, conversion maturity, budget ceiling, and owner. “Find overlap” is not enough; the output must be a reversible action or a documented hold.
Freeze campaign context
Capture campaign type, goals, bidding strategy, budgets, locations, languages, networks, brand settings, Search themes or keywords, negative lists, asset groups, landing pages, feed or audience inputs, conversion actions, and change history. Mark periods around launches, budget changes, tracking migrations, and agency changes. An overlap finding without its goal and measurement context is only a list of shared terms.
Build the inventory map
Map queries or search categories to the campaign, ad or asset group, landing page, conversion action, and commercial intent. Separate brand, nonbrand, competitor, navigational, research, and service-request demand. For each route, record whether the page is useful, available, and owned by the same service boundary.
Use the Performance Max search terms and landing-page views as evidence, but retain their limitations. Some demand is grouped or modeled, and a conversion can occur through inventory that is not explained by a direct search click. Do not treat a shared category as proof that two campaigns delivered the same customer.
Check auction and keyword rules
Confirm whether a Search keyword exactly matches the query, whether match behavior or spelling correction applies, and whether campaign eligibility differs by budget, location, audience, or goal. Google’s interaction guidance says the higher Ad Rank normally wins when campaigns can serve the same impression, with stated exceptions. This is a platform behavior to verify in the current account, not a reason to assume that every impression was transferred.
Record brand exclusions, negative keywords, search themes, and account-level exclusions. An exclusion can protect a route but can also remove useful discovery. Make the intended boundary explicit before changing it.
Compare outcomes, not just conversions
Create a cohort table by campaign, query family, landing page, and outcome maturity:
| Layer | Question | |—|—| | reach | where did eligible demand appear? | | interaction | which route received impressions, clicks, or visits? | | event | which conversion action was recorded? | | quality | was the request accepted and serviceable? | | value | did it create an opportunity, sale, or margin? | | incrementality | what would likely have happened without the route? |
Keep platform-reported conversions separate from CRM and finance outcomes. Compare conversion time, attribution window, consent, duplicate handling, and sales-cycle lag. A campaign that reports more conversions may simply have a broader goal or earlier event.
Audit landing-page continuity
Read the query, asset, landing page, form, confirmation, and first response as one route. A Performance Max asset group that reaches a generic page can appear to overlap with Search while producing a different buyer job. Check fit, scope, proof, geography, response time, capacity, and source context. If the route cannot be served, the correction belongs to the offer or routing as well as the campaign.
Test incrementality carefully
Choose a bounded comparison: a controlled exclusion, matched market, time-boxed holdout, or staged budget change where the account and business can support it. Define primary mature outcome, guardrails for cost, quality, reach, and owner cash, maturity date, effort ceiling, and rollback. Do not make a permanent exclusion from a short period of noisy or immature data.
If an exact-match Search route is strategically important, protect it intentionally and measure whether Performance Max adds incremental nonbrand or cross-channel value. If the campaigns serve different jobs, document the distinction instead of calling normal interaction a defect.
Check goal hierarchy and lag
List every conversion action included in each campaign’s optimisation and reporting columns. Separate an early form event, a qualified lead, an opportunity, and a closed outcome. If Performance Max is trained on a broad event while Search is judged on a later sales stage, their apparent overlap cannot be compared on cost per conversion alone. Use the same mature definition where the data volume and platform setup allow it, and keep earlier events as diagnostic guardrails.
Record the expected delay from click to acceptance and from acceptance to value. Review an early safety window for broken tracking and a later outcome window after the relevant sales cycle. If a campaign has recently changed goals, wait for the cohort to mature before declaring that an exclusion improved performance. A short-term fall in conversions may be a measurement transition; a stable fall in accepted opportunities is a different decision.
Review asset and feed boundaries
Inspect asset groups, final URLs, page feeds, audience signals, product or service labels, and location settings. Two campaigns can appear to overlap in Search while one is intentionally broadening discovery through other Google inventory. Conversely, a generic final URL or duplicated asset group can make the routes indistinguishable to the business even when the auction treats them normally.
Document which campaign is allowed to introduce new demand, which is meant to protect known demand, and which landing pages can accept each job. Give exclusions a reason, owner, date, and rollback test. Revisit them when the offer, brand, market, or conversion goal changes.
Review operational ownership
Name who owns query review, exclusions, landing pages, conversion actions, CRM matching, sales feedback, and budget changes. Keep a decision log that records evidence, hypothesis, action, expected effect, and recheck date. If no one can explain why an exclusion exists or who can roll it back, the governance gap is part of the audit result.
Verdicts
Protect and scale: campaign roles are distinct, measurement is mature, and incremental outcomes support more budget.
Repair measurement: goals, conversion actions, identifiers, or maturity prevent a fair comparison.
Repair route: shared demand reveals a landing, offer, serviceability, or handoff problem.
Test exclusion: overlap is material and a bounded experiment can estimate substitution or incrementality.
Hold: evidence is too sparse, campaign context changed, or rollback is not safe.
The Performance Max Search Overlap Audit Board is complete when it maps campaign context, query and landing routes, auction rules, outcomes, incrementality, exclusions, owners, maturity, and rollback before recommending scale, repair, or pause.
How did this article land?
Choose one reaction. You can change it anytime.
