Performance Max overspending is one of the most common and misunderstood problems in Google Ads today. It happens when PMax campaigns consistently blow through budget allocations while delivering diminishing returns, and the fix is almost never what advertisers try first. This is the story of how a mid-size ecommerce brand burning through roughly $45K per month in Google Ads spend watched its Performance Max ROAS erode for three consecutive months, tried every standard lever Google recommends, made things worse, and ultimately recovered by rebuilding the campaign architecture from the ground up rather than constraining the budget.
The result after 60 days: spend stabilized within 3% of target budgets, ROAS climbed back above pre-decline levels, and revenue from Google Ads grew even though total spend stayed flat. The lesson is transferable to almost any account where PMax is eating budget without accountability. Here is exactly what happened and what got fixed.
The Situation: A Growing Ecommerce Brand With A PMax-Heavy Account
Account Profile And Scale
The brand sells consumer goods across roughly 15 product categories, with average order values ranging from $60 to $200. Google Ads was the primary paid acquisition channel, running about $45K per month across a mix of Performance Max, branded Search, non-branded Search, and Shopping campaigns. The account had been profitable for over a year before the decline started.
PMax accounted for roughly 60% of total spend, which is not unusual for ecommerce accounts that adopted Google's recommendations to consolidate into Performance Max. The brand had an in-house marketing lead managing the account day to day, supplemented by a freelance Google Ads specialist working about 10 hours per week.
The Overspending Pattern
The problem was not that PMax was spending too much in absolute terms. The problem was where that spend was going. Over a three-month window, several patterns emerged simultaneously:
- PMax campaigns consistently spent 15-25% above their daily budget targets
- ROAS dropped from a healthy range to well below breakeven on several asset groups
- Branded search volume did not change, but PMax was claiming an increasing share of branded conversions
- Non-branded Search campaigns saw impression share plummet as PMax cannibalized their traffic
- The Insights tab showed growing spend in Display and YouTube placements with almost no attributable conversions
The brand's marketing lead described it as "watching the budget evaporate into channels we never asked it to run on."
What They Had Already Tried
Before the rebuild, the team had cycled through the standard recommendations: raising target ROAS floors, lowering daily budgets, pausing underperforming asset groups, and adding a few negative keywords at the campaign level. Each adjustment produced a brief improvement followed by a worse relapse. Lowering budgets caused PMax to prioritize cheaper, lower-intent placements. Raising target ROAS made the algorithm chase branded queries more aggressively. Pausing asset groups just concentrated spend into the remaining groups without fixing the underlying issues.
This is the trap most advertisers fall into with Performance Max. The levers Google gives you are blunt instruments that change how much the campaign spends, not where or why. Solving PMax overspending requires structural changes, not budget constraints.
The Diagnosis: Why Performance Max Was Actually Overspending
The root causes were not in the bidding strategy. They were in the architecture of the campaigns themselves.
Asset Group Misalignment Was Inflating Reach
The account had 11 asset groups, loosely organized by product type but with significant overlap in the products, audiences, and landing pages assigned to each group. When asset groups overlap, PMax has no clear signal about which group should serve which query or audience segment. The result is that it spreads spend across all of them inefficiently, often showing the wrong creative for the wrong product to the wrong audience.
URL Expansion Was Sending Spend To Low-Converting Pages
URL expansion was turned on by default across all PMax campaigns. This meant Google was autonomously sending traffic to category pages, blog posts, and even the FAQ section of the site. Some of these pages had zero conversion infrastructure. The brand was paying for clicks that landed on pages with no clear path to purchase.
Audience Signals Were Too Broad Or Missing Entirely
Most asset groups had either no audience signals or extremely broad ones (interests like "shopping enthusiasts"). Without strong audience signals, PMax defaults to Google's own judgment about who to target, which often means chasing the cheapest available impressions rather than the most qualified buyers. This was the primary driver behind the account's growing Display and YouTube spend with negligible returns.
No Account-Level Negative Keywords
The account had a handful of campaign-level negatives but no account-level negative keyword list. This meant PMax was bidding on competitor brand terms, irrelevant product categories, and informational queries that had no purchase intent. The vanity metrics looked acceptable because clicks were coming in, but conversion quality was deteriorating.
PMax And Search Were Fighting Each Other
The non-branded Search campaigns and PMax were competing for the same queries. Google's documentation says PMax takes priority over standard Shopping but defers to exact-match Search keywords. In practice, the interaction is more complicated. The account's broad match Search keywords were losing impressions to PMax, while PMax was winning auctions at higher CPCs for the same queries because it had more budget flexibility. The campaigns were bidding against each other inside the same account.
The Fix: A Structural Rebuild, Not A Budget Cut
The rebuild took about two weeks of active work followed by a careful relearning period. Every change was structural, not just a knob turn on an existing campaign.
Consolidating Asset Groups By Product Category And Margin Tier
The 11 asset groups were collapsed into 5, each aligned to a distinct product category with its own margin profile. High-margin products got their own asset group with more aggressive targets. Lower-margin categories were grouped together with tighter ROAS floors. No overlap in products or landing pages between groups. This gave PMax clear, non-competing signals about what to promote and where.
Restricting URL Expansion To High-Converting Landing Pages
URL expansion was disabled entirely on three of the five asset groups. On the remaining two (where the product catalog was broad enough to benefit from dynamic landing page selection), a final URL expansion filter restricted eligible pages to product detail pages and a shortlist of high-converting collection pages. Blog posts, FAQ pages, and informational content were excluded completely.
Layering Audience Signals From First-Party Data
Each asset group received audience signals built from the brand's actual customer data: purchaser lists segmented by product category, site visitors who had viewed specific product pages, and high-value customer lookalikes. The generic interest-based signals were removed. This did not restrict PMax to those audiences (audience signals are suggestions, not hard targeting), but it gave the algorithm a much stronger starting point for finding qualified buyers.
Building An Account-Level Negative Keyword Foundation
An account-level negative keyword list was built with over 400 terms covering competitor brands, irrelevant product categories, informational intent queries, and known low-converting search terms from the account's historical query data. This was one of the highest-impact changes, immediately reducing spend on queries that had been silently draining budget for months.
Restructuring The PMax And Search Campaign Relationship
Non-branded Search campaigns were rebuilt with exact and phrase match keywords only, eliminating the broad match overlap that had been causing PMax to cannibalize Search traffic. A clear hierarchy was established: Search campaigns owned high-intent, high-converting queries with exact match. PMax handled the broader discovery and Shopping placements. Branded Search was kept separate with its own budget, preventing PMax from inflating its ROAS numbers by claiming branded conversions. This kind of structural scaling work is what separates accounts that grow profitably from accounts that just grow spend.
Results After 60 Days
Spend Stabilization
Within 30 days, PMax campaigns were hitting daily budgets within 3-5% variance, down from the 15-25% overspend pattern before the rebuild. Total account spend stayed flat because the goal was never to cut budget; it was to control where the budget went.
ROAS Recovery
Account-wide ROAS recovered to pre-decline levels within the first 30 days and continued improving through day 60. The high-margin asset groups outperformed their targets consistently. The key shift was not bidding efficiency but placement quality: the same budget was reaching higher-intent audiences on more relevant placements.
Impression Quality Shift
The share of impressions from Display and YouTube placements dropped significantly, replaced by Shopping and Search placements. Branded conversion attribution shifted back to the branded Search campaign where it belonged, which gave the team an honest picture of what PMax was actually contributing.
Revenue Impact
Total revenue from Google Ads grew on flat spend, driven by better conversion rates from higher-quality traffic. The brand's marketing lead described the difference as "finally knowing where the money is going and why."
Lessons: What Actually Controls Performance Max Spending
Performance Max overspending is rarely a budget problem. It is a signal problem. When PMax lacks clear signals about which products to promote, which pages to land on, which audiences to prioritize, and which queries to avoid, it does what any algorithm does with ambiguity: it explores broadly, cheaply, and often wastefully.
The levers that actually control PMax spend are:
- Asset group architecture (what the campaign is structured to promote)
- URL expansion settings (where traffic lands)
- Audience signal quality (who the algorithm starts looking for)
- Negative keywords (what the campaign is not allowed to touch)
- Campaign hierarchy (how PMax interacts with your other campaigns)
Budget caps and target ROAS adjustments are downstream of these structural decisions. If the structure is wrong, no amount of bid adjustment will fix the overspending. The same principle applies to ecommerce accounts running Shopping alongside PMax, where feed optimization and campaign structure do more for profitability than bid management alone.
How The Management Model Changes The Outcome
The reason this account's PMax problem persisted for three months before it was fixed comes down to who was managing the account and how.
Why The Freelancer Approach Failed
The freelance specialist working 10 hours per week was capable but reactive. They saw the symptoms (declining ROAS, overspending) and applied the standard fixes (budget caps, ROAS floors). They did not have the bandwidth or the diagnostic framework to identify the structural causes, rebuild asset groups from scratch, audit URL expansion across every campaign, and construct an account-level negative keyword list from historical query data. This is the ceiling problem with freelancers and even many agencies: execution is capped at whatever one person can physically get through in a week.
How DWY Gives In-House Teams A Diagnostic Framework
For brands with an in-house marketing lead who wants to stay in control, groas's Done With You model pairs the proprietary engine trained on over $500 billion in profitable ad spend with a senior strategist who works alongside your team. The engine runs the heavy analytical work around the clock, identifying structural issues like asset group overlap, URL expansion bleed, and audience signal gaps before they become expensive problems. The strategist provides the diagnostic framework and works through the rebuild with your team, while your people stay in the driver's seat making the final calls. A weekly report shows exactly what changed and why, and strategy calls every other week keep the account moving forward. For brands in this situation, the DWY model would have caught the PMax structural decay within the first two weeks and provided the framework to fix it without the three-month learning-the-hard-way period.
What DFY Execution Looks Like For PMax-Heavy Accounts
For brands that do not want to manage Google Ads at all, groas's Done For You model assigns a dedicated strategist who owns the entire account end to end. The proprietary engine monitors PMax performance continuously, flagging structural drift before it shows up in ROAS numbers. The strategist does not just adjust bids and budgets; they own campaign architecture, asset group design, landing page strategy, and the relationship between PMax and every other campaign in the account. For a PMax-heavy ecommerce account like the one in this story, that means the rebuild happens proactively rather than after three months of declining performance. If you want groas to own your Google Ads as a function, including landing pages, offers, and the full campaign architecture, the right move is to apply for DFY and let the team figure out the right plan on a call.
What This Means For Your Account
If your Performance Max campaigns are consistently overspending, eroding ROAS, or eating into the performance of your Search campaigns, the problem is almost certainly structural, not budgetary. The fix is not a tighter budget cap or a higher ROAS target. It is a rebuild of asset groups, URL expansion settings, audience signals, negative keywords, and campaign hierarchy.
This is work that requires both deep diagnostic capability and the execution bandwidth to implement changes across an entire account simultaneously. It is exactly the kind of problem that separates accounts managed by a single person working limited hours from accounts backed by an engine that never stops analyzing and a strategist who knows where to look first.
groas exists to close that gap. Whether your team wants to stay in control with DWY support or hand off Google Ads entirely with DFY, the proprietary engine and senior strategist work together to prevent the structural decay that causes PMax overspending in the first place, month to month, no long-term contract, with $0 onboarding.
If you have an in-house team and want better tooling plus senior advisory, get started with DWY. If you want groas to own your Google Ads end to end, apply for DFY and the team will determine the right fit on a call.
Frequently Asked Questions About Performance Max Overspending
Why Is My Performance Max Campaign Overspending Its Daily Budget?
Performance Max can spend up to 2x your daily budget on any given day, which Google considers normal behavior. But consistent overspending beyond that threshold usually signals a structural problem, not a bidding issue. The most common causes are asset group overlap (where multiple groups compete for the same traffic), URL expansion sending clicks to low-converting pages, and weak or missing audience signals that cause the algorithm to chase cheap, low-intent placements. Fixing PMax overspending requires rebuilding campaign architecture, not just lowering budget caps or raising ROAS targets. Budget constraints only change how much PMax spends, not where or why it spends.
Does Lowering My Performance Max Budget Fix Overspending?
No. Lowering your PMax budget often makes things worse. When you constrain the budget without fixing the underlying structure, the algorithm compensates by shifting spend toward cheaper placements, typically Display and YouTube inventory with lower intent and lower conversion rates. This can actually accelerate ROAS decline. The effective fix is structural: consolidate asset groups, restrict URL expansion, strengthen audience signals, and build a robust negative keyword foundation. These changes control where spend goes, which is the real problem behind PMax overspending.
How Do I Stop Performance Max From Cannibalizing My Search Campaigns?
The key is establishing a clear campaign hierarchy. Rebuild your non-branded Search campaigns with exact and phrase match keywords only, removing broad match overlap that PMax exploits. Keep branded Search in a separate campaign with its own budget so PMax cannot inflate its numbers by claiming branded conversions. This gives Search priority on high-intent queries while PMax handles broader discovery and Shopping placements. Without this structure, PMax and Search bid against each other inside your own account, driving up CPCs.
Should I Turn Off URL Expansion In Performance Max?
In most cases, yes, or at minimum restrict it heavily. URL expansion lets Google send traffic to any page on your site, including blog posts, FAQ pages, and category pages with no conversion path. For ecommerce accounts, restrict eligible URLs to product detail pages and a shortlist of high-converting collection pages. Disable it entirely on asset groups where you need tight control over the landing experience. This single change can dramatically improve conversion rates by ensuring every click lands on a page designed to convert.
What Are The Best Audience Signals For Performance Max?
The strongest audience signals come from your own first-party data: purchaser lists segmented by product category, site visitors who viewed specific product pages, and high-value customer lookalikes. Generic interest-based signals like "shopping enthusiasts" give PMax almost no useful direction. Remember that audience signals in PMax are suggestions, not hard targeting. The algorithm will expand beyond them. But strong first-party signals give it a much better starting point for finding qualified buyers rather than chasing the cheapest available impressions.
How Long Does It Take To Fix A Performance Max Overspending Problem?
A full structural rebuild typically takes one to two weeks of active work, followed by a two to four week relearning period as the algorithm adjusts to the new architecture. In the account described in this article, spend stabilized within 30 days and ROAS recovered to pre-decline levels by day 60. The timeline depends on account complexity, but the critical point is that results come from structural changes, not incremental bid adjustments.
Can groas Fix My Performance Max Overspending Problem?
Yes. groas is purpose-built for exactly this kind of structural problem. The proprietary engine trained on over $500 billion in profitable ad spend identifies issues like asset group overlap, URL expansion bleed, and audience signal gaps continuously, not just when a human has time to look. With the Done With You model, a senior strategist works alongside your in-house team to diagnose and rebuild the account while your people stay in control. With Done For You, a dedicated strategist owns the entire account end to end and prevents structural decay before it shows up in your ROAS numbers. Both are month to month with $0 onboarding.
Why Do Standard Google Recommendations Make PMax Overspending Worse?
Google's standard recommendations, such as raising target ROAS floors or lowering daily budgets, treat symptoms rather than causes. Raising target ROAS pushes the algorithm toward branded queries where it can hit the target easily, masking the real problem. Lowering budgets forces PMax into cheaper, lower-quality placements. These are blunt instruments that change spending volume without addressing where or why the budget is being wasted. Effective PMax management requires structural diagnosis that goes beyond what Google's in-platform recommendations offer.
Is It Better To Use DWY Or DFY From groas For PMax-Heavy Accounts?
It depends on whether you have someone in-house who knows Google Ads and wants to stay involved. Done With You is ideal if your team wants to keep running the account day to day with better tooling and senior advisory from groas. Done For You is the right fit if you want groas to own Google Ads as a function, including campaign architecture, landing pages, and the full PMax rebuild. If you are unsure, apply for DFY and the groas team will determine the right plan on a call. Many brands start with DWY and upgrade to DFY as they scale.
How Do I Know If My Performance Max Account Needs A Structural Rebuild?
Look for these signals: PMax consistently exceeding daily budgets by more than 10%, ROAS declining over multiple weeks despite bid adjustments, growing spend on Display and YouTube with few attributable conversions, non-branded Search impression share dropping while PMax spend rises, and PMax claiming an increasing share of branded conversions. If you have tried budget caps and target ROAS changes without lasting improvement, the problem is structural. An account audit focused on asset group overlap, URL expansion settings, audience signals, and campaign hierarchy will confirm whether a rebuild is needed.
.png)
