Performance Max cannibalizing Search campaigns is one of the most common structural problems in ecommerce Google Ads accounts, and one of the hardest to spot when you are only looking at campaign-level ROAS. PMax Search cannibalization occurs when Performance Max quietly absorbs impression share, branded traffic, and high-intent queries that would otherwise convert through dedicated Search campaigns, inflating PMax results while suppressing overall revenue growth. This is a walkthrough of how one ecommerce brand diagnosed and fixed exactly that problem, recovering meaningful revenue within 90 days by restructuring how PMax and Search coexisted in the same account.
The brand in this case was spending in the range of $80K per month across Google Ads, running a mix of Performance Max, branded Search, non-branded Search, and Shopping campaigns. On paper, the account looked healthy. In practice, it had been stuck at the same revenue ceiling for nearly five months, and every attempt to scale spend just raised costs without moving the needle. Here is what was actually happening and what got fixed.
The Setup: A High-Spend Ecommerce Brand Stuck At The Same ROAS
Account Profile And Campaign Mix
This was a mid-market ecommerce brand selling a broad catalog across multiple product categories. Monthly ad spend sat around $80K, split roughly 45% into Performance Max, 30% into non-branded Search, 15% into branded Search, and 10% into standalone Shopping. Attribution was set to last-click at the time, with Google Analytics running alongside Google Ads conversion tracking.
The account had been managed by an in-house team of two: a performance marketer handling day-to-day optimizations and a marketing director reviewing weekly reports. They had launched Performance Max about eight months earlier, consolidating several older Shopping and Display campaigns into three broad PMax campaigns with multiple asset groups each.
What "Stuck" Looked Like In The Data
The surface metrics told a reassuring story. Overall account ROAS hovered around 5x. PMax ROAS was consistently above 6x. The team was hitting their efficiency targets. But revenue had flatlined. Conversion volume was not growing despite incremental budget increases. CPCs on non-branded Search had climbed steadily over the prior quarter. And when you plotted total revenue against total spend on a monthly basis, the curve was flat, not scaling.
The instinct was to push more budget into PMax, since it was reporting the highest ROAS. That instinct was exactly the wrong move, and this is a pattern that shows up frequently when ROAS looks strong but revenue stays flat.
The Diagnosis: Where The Budget Was Actually Going
How Performance Max Was Cannibalizing Top-Performing Search Campaigns
The first signal appeared in the Search impression share data. Over the prior three months, the non-branded Search campaigns had lost significant impression share to budget, even though total account budget had increased. Meanwhile, PMax was serving on an expanding set of search terms, many of which overlapped directly with the non-branded Search campaigns' highest-converting keywords.
By pulling the PMax insights report and cross-referencing search term categories against the existing Search campaign structure, the overlap became obvious. PMax was bidding on and winning auctions for queries that were already covered by well-optimized exact and phrase match keywords in dedicated Search campaigns. Because PMax operates as a single campaign type across all Google inventory, it was absorbing budget that could have gone to Search campaigns with tighter controls and better conversion rates for those specific queries.
What The Placement Report Revealed
A deeper look at where PMax impressions were actually serving told the rest of the story. A substantial share of PMax impressions were going to Display Network placements and YouTube inventory with low conversion rates. PMax was reporting strong aggregate ROAS because it was also capturing high-intent branded and non-branded search traffic, which carried the overall number. The low-quality Display and YouTube placements were being subsidized by the search traffic PMax was absorbing.
This is the core mechanism behind PMax cannibalizing Search campaigns: PMax looks efficient because it takes credit for conversions that Search campaigns would have captured. But it spends that absorbed budget partly on low-quality inventory that does not convert, meaning total revenue per dollar goes down even while PMax ROAS looks fine.
The Attribution Blind Spot
Under last-click attribution, PMax was getting full credit for conversions where it served the final ad. But many of those conversions involved earlier touchpoints from Search campaigns that had been doing the actual demand capture work. The result: PMax looked like the hero, Search looked like it was declining in value, and the logical response (shift budget to PMax) made the problem worse. This is a textbook example of why high ROAS numbers can lie when automation runs without proper structural oversight.
The Fix: Restructuring Budget Allocation Between PMax And Search
Separating Brand, Search, And PMax Budgets
The first structural change was establishing hard budget boundaries. The team created a dedicated branded Search campaign with its own budget, isolated from PMax entirely. Non-branded Search campaigns were given a protected budget allocation that could not be cannibalized by PMax's automated bidding. PMax was reduced to roughly 30% of total spend, down from 45%, and given a clearly defined role: prospecting through Shopping, Display, and YouTube, not competing for high-intent search queries.
This separation is the single most important step in preventing PMax from overspending at the expense of Search. Without hard budget walls, Google's algorithm will naturally shift spend toward PMax because it has more inventory options, regardless of whether that shift actually helps revenue.
The Asset Group Rebuild
The three broad PMax campaigns with multiple asset groups each were collapsed into a tighter structure. Instead of organizing asset groups by product category with overlapping audience signals, the team rebuilt around distinct audience themes: prospecting cold audiences, re-engaging site visitors, and targeting high-value customer segments. Each asset group got a focused set of creatives and signals rather than the kitchen-sink approach.
Fewer asset groups with stronger signals gave the algorithm less room to drift into Search territory and more reason to focus on genuine prospecting, which is where PMax adds value that Search and Shopping cannot replicate on their own. For brands running broad catalogs, tightening the Shopping feed itself is another lever that compounds these structural fixes.
Applying Negative Keyword Lists To Protect Search From PMax Overlap
Account-level negative keyword lists were applied to the PMax campaigns to exclude branded terms and the top-performing non-branded queries already covered by Search. This is one of the most effective PMax budget protection mechanisms available, and it is still underused. By blocking PMax from bidding on terms where Search campaigns already had strong quality scores and conversion history, the team ensured that each campaign type served its intended role.
The combination of budget separation, asset group restructuring, and negative keyword application created a clean architecture where PMax and Search could coexist without competing for the same traffic.
The Result: What Changed In 90 Days
Within the first 30 days, the changes were already visible. Non-branded Search impression share recovered, and the CPCs on those campaigns dropped as they no longer competed against the brand's own PMax campaigns in the same auctions. By day 60, total conversion volume had increased meaningfully while overall spend stayed within a few percentage points of the pre-restructure level.
By the 90-day mark, the trajectory had shifted. Revenue was growing again. The non-branded Search campaigns were driving a larger share of total conversions at a lower cost per acquisition. PMax was still contributing, but its role had shifted to genuine prospecting and upper-funnel activity rather than absorbing Search demand. Overall account ROAS stayed within a competitive range while revenue climbed, which is the outcome that matters. A high ROAS number on a flat revenue line is not growth; it is a ceiling.
The campaigns that drove the recovery were the non-branded Search campaigns that had been starved of budget and the restructured PMax campaigns that were now focused on incremental audiences rather than cannibalizing existing demand.
The Lesson: What This Brand's PMax Setup Was Getting Wrong (And What Most Accounts Get Wrong Too)
Five Structural Mistakes That Cause PMax To Cannibalize Search
-
No budget separation between branded Search, non-branded Search, and PMax. Without hard budget walls, PMax absorbs high-intent traffic first because it can.
-
Broad asset groups with overlapping audience signals. This gives PMax's algorithm permission to bid on everything, including queries your Search campaigns already own.
-
No negative keyword lists on PMax campaigns. Without explicit exclusions, PMax will compete with your own Search campaigns in the same auctions.
-
Evaluating PMax on campaign-level ROAS without checking for overlap. A 6x ROAS on PMax means nothing if it is just taking credit for traffic Search would have converted at 8x.
-
Defaulting to last-click attribution without cross-campaign analysis. Last-click rewards whichever campaign served last, not whichever campaign actually influenced the purchase.
How To Spot The Same Problem In Your Own Account
Check your non-branded Search impression share trends over the past 90 days. If impression share lost to budget is rising while your total account budget is flat or growing, PMax may be absorbing your Search budget. Pull the PMax search term insights and compare them against your Search campaign keywords. If there is significant overlap on your highest-converting terms, you have a cannibalization problem. Finally, look at total revenue growth, not campaign-level ROAS. If ROAS looks fine but revenue is stuck, the attribution layer is hiding the real issue.
What Fully Managed Execution Catches That In-House Teams Often Miss
The challenge with PMax and Search coexistence is that it requires constant structural vigilance. It is not a one-time fix. As PMax's algorithm evolves, as your product catalog changes, as competitive dynamics shift, the boundaries between campaign types need active management. An in-house team with two people and a full plate of responsibilities beyond Google Ads rarely has the bandwidth to run campaign interaction analysis on a weekly basis.
This is where groas changes the math. With groas DFY, a dedicated strategist owns the entire account end-to-end, backed by a proprietary engine trained on over $500 billion in profitable ad spend. Campaign interaction analysis, budget protection between PMax and Search, negative keyword maintenance, asset group optimization, and attribution modeling are not occasional audits. They are part of the ongoing management cadence. The engine monitors overlap patterns continuously, and the strategist restructures before cannibalization has time to suppress revenue.
The brand in this case study had the data available to diagnose the problem. What they did not have was the bandwidth or the pattern recognition to catch it before five months of flat revenue had already passed. A groas strategist would have flagged the cannibalization signal within the first few weeks, because they have seen this exact structural pattern play out across hundreds of accounts.
For in-house teams that want to stay in control but need that level of pattern recognition on their side, groas DWY puts the engine and a senior strategist alongside your team while you keep driving day-to-day decisions. For brands that want to hand over execution entirely and not worry about whether PMax is quietly eating their Search budget, DFY is the path.
Where This Leaves You
PMax is not inherently broken. But it is structurally incentivized to absorb as much budget as Google's algorithm decides it should, and that decision does not always align with your revenue goals. The fix is architectural: budget walls, negative keyword protection, tighter asset groups, and continuous monitoring of campaign interaction patterns.
Most accounts will not catch this problem until revenue has already stalled, because the surface-level metrics look healthy. The brands that avoid the plateau are the ones that have someone watching the structural layer constantly, not just the performance layer.
If your ecommerce account is running PMax alongside Search and revenue has stopped growing despite stable or rising ROAS, the overlap is the first place to look. If you want someone to look for you, with the engine and the expertise to fix it before another quarter goes flat, apply for groas DFY and let the team diagnose it on the call. No onboarding fees, no long-term contracts, cancel anytime. groas earns the next month by performing.
Frequently Asked Questions
How Do I Know If Performance Max Is Cannibalizing My Search Campaigns?
Check your non-branded Search impression share trends over the past 90 days. If impression share lost to budget is rising while your total account budget is stable or growing, PMax is likely absorbing your Search budget. Cross-reference the PMax search term insights report against your Search campaign keywords. Significant overlap on your highest-converting terms confirms cannibalization. Also compare total revenue growth against campaign-level ROAS. If ROAS looks healthy but revenue is flat, the problem is structural. groas DFY catches this pattern early because a dedicated strategist monitors campaign interaction data continuously, backed by an engine trained on over $500 billion in profitable ad spend.
How Do I Prevent PMax From Overspending On Search Queries?
The most effective PMax budget protection mechanisms are hard budget separation between branded Search, non-branded Search, and PMax campaigns, combined with negative keyword lists applied at the PMax campaign level to exclude your top-performing Search terms. You also need to tighten asset groups so PMax's algorithm has less room to drift into Search territory. Budget walls prevent the algorithm from reallocating spend away from high-performing Search campaigns automatically. Without these protections, PMax will naturally absorb high-intent traffic because it has access to more inventory options.
Can You Use Negative Keywords In Performance Max Campaigns?
Yes. Google now allows account-level negative keyword lists to be applied to Performance Max campaigns. This is one of the most effective ways to prevent PMax from bidding on branded terms and high-converting non-branded queries that your Search campaigns already cover well. Apply exclusions for your brand terms and your top-converting exact match keywords. This forces PMax to focus on genuine prospecting rather than absorbing demand your Search campaigns would convert more efficiently.
What Is The Best Budget Split Between Performance Max And Search?
There is no universal ratio because it depends on your catalog size, competitive landscape, and where your revenue actually comes from. A practical starting point for ecommerce accounts is to give Search campaigns a protected budget allocation covering your highest-intent, highest-converting queries, and assign PMax a defined prospecting budget. The brand in this case study moved PMax from 45% to roughly 30% of total spend and saw revenue grow. The key principle is that PMax should not compete with Search for traffic your Search campaigns already convert well.
Why Does Performance Max Report High ROAS But Revenue Stays Flat?
PMax reports high ROAS because it absorbs high-intent search traffic, including branded queries, that converts easily and inflates its aggregate numbers. Meanwhile, it also spends budget on low-conversion Display and YouTube placements. The high-intent traffic subsidizes the low-quality placements in the ROAS calculation. Because PMax is taking credit for conversions Search would have captured, and spending part of that budget on inventory that does not convert, total revenue per dollar actually declines even though PMax ROAS looks strong.
Should I Turn Off Performance Max Entirely?
Not necessarily. PMax adds genuine value for prospecting through Shopping, Display, and YouTube inventory when it is properly structured. The issue is not PMax itself but how it interacts with your Search campaigns when boundaries are not enforced. The fix is architectural: separate budgets, negative keyword exclusions, tighter asset groups, and ongoing monitoring. Turning PMax off entirely means losing its prospecting capabilities. Restructuring it to stay in its lane means keeping the upside without the cannibalization.
How Long Does It Take To Fix PMax Search Cannibalization?
The structural changes (budget separation, negative keyword application, asset group rebuild) can be implemented within a few days. Results typically begin showing within the first 30 days as Search impression share recovers and CPCs normalize. Full revenue recovery and a clear upward trajectory usually take 60 to 90 days as the algorithm adjusts and the restructured campaigns accumulate conversion data in their new roles.
How Does groas Handle PMax And Search Coexistence Differently?
With groas DFY, campaign interaction analysis is not an occasional audit. It is part of the ongoing management cadence. A dedicated strategist owns the entire account, backed by a proprietary engine that monitors overlap patterns, budget allocation drift, and attribution discrepancies continuously. The strategist restructures before cannibalization suppresses revenue, rather than reacting after months of flat performance. This combination of constant engine monitoring and senior human oversight is what prevents the five-month revenue plateaus that in-house teams typically experience before they even identify the root cause.
Does Last-Click Attribution Make PMax Cannibalization Worse?
Yes. Last-click attribution rewards whichever campaign served the final ad before conversion, regardless of which campaign actually captured the demand. When PMax serves a final impression on a user who was already driven by a Search ad, PMax gets full credit. This makes PMax look more valuable and Search look less valuable, which leads teams to shift more budget into PMax, accelerating the cannibalization cycle. Cross-campaign analysis and data-driven attribution models help reveal the true contribution of each campaign type.
.png)
