Since their widespread deployment, Performance Max campaigns have generated as much enthusiasm as skepticism. Promised as a global lever capable of activating the entire Google inventory, they are quickly establishing themselves in Google Ads accounts. However, one concern consistently arises among advertisers: the gradual decline in classic Search performance after the activation of Performance Max.
Behind this impression lies a much more subtle advertising mechanism than a simple budget transfer.
Silent Redistribution of Already Profitable Queries
Performance Max relies on automation and all available signals in an account. This includes queries already known to generate conversions via Search.
When Performance Max is activated, Google can redirect certain impressions to this new format, even if a Search campaign was already targeting these keywords.
In reality, it is not a disappearance of queries, but a change in attribution channel.
Analyses conducted by several agencies show that on average 15 to 30% of conversions initially attributed to Search shift to Performance Max in the first few weeks. The overall volume remains stable, but the account’s reading changes radically.
Algorithmic Priority Given to Automated Formats
Google increasingly favors campaigns capable of using its proprietary signals on a large scale. Performance Max checks all the boxes: audience, intent, history, multi-device behaviors.
Compared to a classic Search campaign, even a well-structured one, Performance Max often benefits from an implicit display priority, especially on broad or ambiguous queries.
Search does not disappear, but it loses visibility in certain areas of the SERP.
This phenomenon is particularly visible in e-commerce accounts, where generic transactional queries gradually shift to Performance Max.
Illusion of Loss Linked to Reporting Change
One of the main sources of confusion comes from reporting.
Before Performance Max, Search concentrated a large part of visible conversions. After activation, these conversions are redistributed across several channels aggregated into a single campaign.
Result: Search seems to decline, while the overall volume remains similar.
According to an internal study conducted by WordStream, nearly 40% of advertisers mistakenly interpret this redistribution as a real decline, due to not comparing data at the account level rather than by campaign type.
The feeling of cannibalization often arises from a reading bias rather than an actual degradation.
Perceived Loss of Control Over Historical Queries
Classic Search is based on an explicit logic: keywords, ads, landing pages.
Performance Max operates in the opposite way, with an automatic assembly logic and very limited visibility on the exact queries triggered.
For advertisers accustomed to finely managing their campaigns, this opacity gives the impression that Performance Max “steals” queries that have been controlled for years.
In reality, Google does not remove these queries, it reintegrates them into a multi-channel logic, where the main objective becomes overall conversion rather than the performance of an isolated lever.
Amplified Effect on Strongly Branded Accounts
Brands with a high volume of searches related to their name are particularly affected.
Performance Max easily captures these queries, often at low cost and with a high conversion rate.
In this context, the brand Search can show a rapid decline in clicks and conversions.
Observed data shows that up to 50% of branded traffic can be absorbed by Performance Max if no exclusion is implemented.
This situation strongly fuels the idea of cannibalization, while it is mainly an algorithmic arbitration in favor of the most automated format.
Internal Competition on Intent Signals
Performance Max and Search sometimes exploit the same user intent signals.
When both campaigns target similar audiences, Google dynamically chooses the one it deems most likely to achieve the defined goal.
This internal competition is not directly visible, but it results in sudden variations in Search distribution, without apparent parameter changes.
This behavior is common in accounts where Performance Max has a more recent or larger conversion history than some Search campaigns.
Misreading of Cost per Conversion
After activating Performance Max, it is common to see an increase in cost per conversion on Search.
This signal is often interpreted as a degradation of profitability.
In reality, Performance Max primarily absorbs the easiest conversions. Search then ends up with more competitive queries or higher in the funnel.
The cost increases, but the role of Search evolves.
According to several e-commerce account audits, Search retains a strong assistance value, even when its direct ROAS decreases after the arrival of Performance Max.
Progressive Stabilization After the Learning Phase
The first weeks following the launch of Performance Max are rarely representative.
The algorithm tests, redistributes, adjusts the activated channels.
During this phase, Search can show significant, sometimes abrupt variations.
In many observed cases, stabilization appears after 4 to 6 weeks, with a more coherent distribution between campaigns.
Accounts that evaluate the situation too early often make unnecessary corrective decisions, artificially reinforcing the sensation of conflict between levers.
Global Arbitration Rather Than Real Cannibalization
In the majority of analyzed configurations, Performance Max does not destroy Search.
It repositions it within a broader advertising architecture, oriented towards global conversion rather than channel-driven management.
The decline of Search is not systematic, and when it exists, it often corresponds to a logical redistribution of the most profitable opportunities.
Advertisers who analyze their results at the account level frequently find that the total volume of conversions increases, even if the reading by lever becomes less flattering.
Advanced Reading to Avoid False Conclusions
Interpreting Performance Max as a direct competitor to Search is a reductive view.
It is rather a change in internal hierarchy, where Google favors campaigns capable of covering the entire user journey.
The sensation of cannibalization appears mainly when reasoning in silos.
A global reading, integrating the complementarity of signals and the redistribution of roles, allows avoiding erroneous diagnoses and hasty decisions.