Performance Max & Multi-Platform Ads for Online Stores
by Character Strategy
Ask ten e-commerce advertisers about Performance Max and you'll start a fight. Half swear it prints money; half swear it quietly ate their brand traffic and called it growth. Both camps are right, which is exactly the problem. PMax is Google's most automated campaign type, and automation is only as trustworthy as the structure wrapped around it.
We run PMax for online stores with the guardrails that make its numbers mean something: brand exclusions so it can't claim customers who were already coming, asset group structure that mirrors your catalog, and incrementality checks that ask the only question that matters: are these sales new, or relabeled? Because we manage Meta and TikTok alongside Google, we have no incentive to defend PMax when the budget would work harder elsewhere. Sometimes it would, and we say so.
Challenges facing Performance Max for E-commerce advertisers.
No visibility into where budget goes
PMax bundles Search, Shopping, YouTube, Display, and Gmail into one campaign with limited placement reporting, so you can't easily see whether spend went to high-intent shoppers or cheap Display filler.
Cannibalizing your other campaigns
Without brand exclusions, PMax serves on your own brand searches and absorbs traffic from your Shopping and Search campaigns, then reports the shifted revenue as its own success.
Automation without enough signal
PMax needs conversion volume to learn. Launched on a low-data account with no audience signals or exclusions, the algorithm optimizes toward whatever converts easiest, usually people you already own.
How we solve these problems.
Asset groups that mirror your catalog
We structure PMax with dedicated asset groups per product category, each with tailored creative, landing pages, and audience signals matched to that buyer.
Brand and placement exclusions
We exclude brand terms and junk placements before launch, so PMax has to earn its results from new demand instead of relabeling traffic you already had.
Incrementality monitoring
We check whether PMax drives sales that wouldn't have happened anyway, using holdouts and channel-level analysis rather than taking the campaign's word for it.
What Performance Max genuinely does well
Credit where due: for e-commerce specifically, PMax has real strengths. It reaches across Google's entire inventory, Search, Shopping, YouTube, Display, Gmail, Discover, from a single campaign, which means it can find converting demand in corners you'd never build dedicated campaigns for. It's particularly good at long-tail query coverage: with a large catalog and a clean feed, the automation matches products to obscure searches at a scale no human account manager could replicate.
It works best under specific conditions: meaningful conversion volume for the algorithm to learn from, accurate conversion tracking feeding it (garbage signal in, garbage optimization out), a well-optimized product feed, and margin room to tolerate the exploration spend automation needs. Stores that meet those conditions and add proper guardrails often get genuinely incremental growth from it. Stores that don't meet them get an expensive black box, which is where the horror stories come from.
The black box problem and brand cannibalization
PMax's biggest cost isn't wasted clicks; it's wasted confidence. Reporting is deliberately coarse: you get limited visibility into which searches triggered your ads, only partial insight into how spend split across channels, and Google has added controls slowly and reluctantly. When a PMax campaign reports a 6x ROAS, the honest response is a question: six times return on what kind of traffic?
Too often the answer is: your own. Left unexcluded, PMax happily serves on searches for your brand name, buyers who had already decided, and books their purchases as campaign wins. It can also absorb traffic your Standard Shopping campaigns were serving. The revenue is real; the incrementality isn't. This is why accounts often "collapse" when brand terms finally get excluded: the reported ROAS drops to what the campaign was genuinely earning all along.
Our standard operating procedure: brand exclusions applied at launch, not after suspicion sets in; channel-split analysis using the reporting that does exist plus scripts that extract more; and periodic holdout thinking, comparing total account performance with PMax scaled up versus down, because campaign-level attribution can't answer the incrementality question by itself.
Feed-only PMax, full asset groups, and when to skip it entirely
There are two quite different ways to run PMax for a store. Feed-only, where you supply the product feed and no creative assets, constrains serving toward Shopping-style placements, behaving something like the old Smart Shopping campaigns. It's the more controlled, more measurable flavor, and often the right starting point for stores whose profit lives in high-intent product searches. Full asset groups, loaded with images, video, and copy, open up Display, YouTube, and Discover reach: more scale, more top-of-funnel exposure, and more places for budget to go quietly underperform if your feed and signals aren't strong.
Just as important is knowing when not to use PMax at all. Low conversion volume starves the algorithm and produces erratic results. Thin margins can't fund the exploration automation requires. Accounts still fixing feed quality or conversion tracking should fix those first, since PMax amplifies whatever data it's fed. And if Standard Shopping is already delivering with control you value, the burden of proof sits on PMax to beat it in a fair test, not on you to adopt it because the interface recommends it. We run that test properly, and on performance terms: if we don't improve your results, you don't pay.
Common questions about performance max for e-commerce.
It can be genuinely strong under the right conditions: solid conversion volume, accurate tracking, an optimized product feed, and guardrails like brand exclusions in place from day one. Under those conditions, PMax finds long-tail demand across Google's whole network at a scale manual campaigns can't match. Without them, it tends to optimize toward your easiest conversions, often your own brand traffic, and report the result as growth. The campaign type isn't good or bad; the setup decides which experience you get.
By default, yes, it's eligible to serve on searches for your own brand, and those clicks convert so easily that the algorithm gravitates toward them. The result is inflated ROAS built partly on customers who were coming anyway. The fix is brand exclusions, which Google now supports through brand lists applied to the campaign. Expect reported performance to drop when you add them. That drop isn't damage; it's the honest number emerging from underneath the flattering one.
A PMax campaign built with your product feed but no uploaded creative assets. Without images, video, and copy to distribute, serving concentrates on Shopping-style product placements rather than Display and YouTube, making it behave somewhat like the retired Smart Shopping format. Stores use it to get PMax's bidding automation on high-intent product searches while limiting budget leakage into low-intent inventory. It's often the sensible first version of PMax for a store, with full asset groups added later if the economics support broader reach.
Think in weeks, not days, and the honest variable is conversion volume: a store feeding the algorithm hundreds of conversions a month stabilizes much faster than one feeding it twenty. Early performance is typically erratic while the system explores. The discipline that matters is not resetting the clock: every significant change to targets, budgets, or structure sends the campaign back into learning. We make changes in measured steps and judge performance on multi-week windows, not on any single alarming Tuesday.
Only partially, and it's fair to be annoyed about that. Google provides asset group and product-level performance, search themes and insights reports, and gradually improved channel reporting, but nothing close to the query-level and placement-level transparency of Standard Shopping and Search. We extract more using scripts that estimate channel splits and by reading product-level data against the rest of the account. The honest position: you manage PMax by controlling inputs, feed, exclusions, signals, and verifying outputs at the account level, not by auditing every placement.
Skip it, or postpone it, when conversion volume is too low for the algorithm to learn reliably, when margins are too thin to fund its exploration spend, or when your feed and conversion tracking still have known problems, because PMax amplifies bad data rather than forgiving it. It's also reasonable to stay on Standard Shopping when it's performing and you value query-level control. PMax should have to beat your current setup in a fair, brand-excluded test before it earns the budget.
Real results in e-commerce.
Browse our e-commerce case studies to see what we can do.
1000+ Conversions in Month One
Digital Marketplace Startup
Challenge
A new digital marketplace needed to prove traction to investors. They had zero brand awareness and a tight timeline to hit conversion targets.
1K+
First Month Conversions
$6.93
Cost Per Conversion
5.2x
ROAS
Results
- Over 1,000 conversions in the first month
- Achieved a $6.93 cost per conversion
- Provided investor-ready performance dashboards
“Over 1000 conversions the first month at a $6.93 CPA. Unbelievable efficiency for a new launch.”
- Rachel S., CEO, Digital Marketplace
Google Shopping ROAS Hits 8.4x
Premium Home Goods Brand
Challenge
A direct-to-consumer home goods brand had stagnant Shopping campaigns with a 2.1x ROAS. They needed to scale profitably without diluting margins.
8.4x
ROAS
up from 2.1x
+210%
Revenue Growth
+35%
CTR Improvement
Results
- ROAS improved from 2.1x to 8.4x over 4 months
- Revenue increased 210% while spend only increased 60%
- Product feed optimization lifted click-through rate by 35%
Seasonal Campaign Drives Record Q4
Specialty Gift Retailer
Challenge
A gift retailer generated 60% of annual revenue in Q4 but had never run a structured seasonal campaign. Previous years saw wasted spend and stockout issues.
+85%
Q4 Revenue
vs prior year
6.7x
BFCM ROAS
0
Stockouts
Results
- Q4 revenue exceeded previous year by 85%
- Black Friday/Cyber Monday ROAS hit 6.7x
- Zero stockout incidents thanks to demand forecasting alignment
Apparel Brand ROAS Climbs from 2.1 to 3.4
DTC Apparel Brand
Challenge
A mid-market apparel brand was bleeding spend on broad discovery traffic. ROAS sat at 2.1 with no clear path to improvement.
3.4x
ROAS
from 2.1x
$200K
Monthly Revenue
+60%
+62%
ROI Improvement
Results
- ROAS climbed from 2.1 to 3.4 without increasing spend
- Revenue increased while waste dropped
- Brand scaled without relying on constant creative refreshes
Supplements Brand CPA Down 35%
DTC Supplements Brand
Challenge
A supplements brand in a competitive category needed to lower acquisition costs without sacrificing volume on the same budget.
$48
CPA
-35%
4.8x
ROAS
from 4.0x
$190K
Monthly Revenue
Results
- CPA dropped 20-40% while total sales increased
- Brand stopped paying for education traffic
- Spend focused on buyers ready to convert
Furniture Brand Scales 10x, Holds 7x ROAS
High-AOV Furniture Retailer
Challenge
A high-AOV furniture retailer needed to scale spend aggressively without destroying efficiency, avoiding the common scale cliff seen in furniture e-commerce.
$210K
Monthly Revenue
from $20K
7.0x
ROAS
held
10x
Spend Scaled
Results
- Spend scaled 10x in four months while holding 7x ROAS
- Profitability remained intact throughout scaling
- Brand avoided the common scale cliff
“They took us from $3K to $30K in spend and revenue followed. We never thought we could scale that fast without losing margin.”
- Amanda P., Founder & CEO
Home Decor Shopping Revenue Up 60%
Home Decor Store
Challenge
A home decor store with 400+ SKUs needed to unlock growth from Google Shopping without increasing spend. The product feed was messy and low-margin items ate budget.
4.6x
ROAS
from 3.2x
+60%
Shopping Revenue
400+
SKUs Optimized
Results
- Shopping revenue increased 40-80% with the same budget
- More spend flowed to profitable products
- Overall account efficiency improved
Electronics Revenue Doubles, ROAS Hits 3.2
Electronics Retailer
Challenge
PMAX was over-spending on low-intent traffic in thin-margin categories. Revenue was stuck at $180K/month with a 1.8x ROAS.
3.2x
ROAS
from 1.8x
$420K
Monthly Revenue
+133%
$115K
Monthly Spend
Results
- ROAS rose from 1.8 to 3.2 within 90 days
- Revenue more than doubled to $420K/month
- Waste fell and scaling became controllable
Outdoor Gear Revenue Grows 60% YoY
Outdoor Gear E-commerce Brand
Challenge
An outdoor gear brand needed year-over-year growth without leaning on constant promotions. How-to and hobby-research traffic was eating budget.
$670K
Annual Revenue
+60% YoY
4.4x
ROAS
from 3.8x
$135
AOV
Results
- Google Ads revenue grew 60% year-over-year
- ROAS improved from 3.8 to 4.4
- Steadier demand outside peak weekends
Skincare Brand Scales 5x, Holds 4.5x ROAS
DTC Skincare Brand
Challenge
A skincare brand with repeat-purchase economics needed to scale spend 5x without ROAS collapse in a crowded category.
$225K
Monthly Revenue
from $45K
4.5x
ROAS
held
$25
CPA
held
Results
- Spend scaled from $10K to $50K/month while maintaining 4.5x ROAS
- Revenue grew predictably from $45K to $225K/month
- Acquisition costs stayed stable enabling inventory investment
“We went from $10K to $50K in spend and our ROAS never dipped. That gave us the confidence to invest in inventory and retention.”
- Lisa M., Head of Growth
Accessories Brand ROAS Jumps from 2x to 5x
DTC Accessories Brand
Challenge
A DTC accessories brand was relying on broad traffic with a 2.0x ROAS. Profitability was declining and scaling felt impossible.
5.0x
ROAS
from 2.0x
$150K
Monthly Revenue
+150%
$20
CPA
-38%
Results
- ROAS increased from 2.0 to 5.0 and revenue doubled
- CPA dropped from $32 to $18-22
- Scaling became repeatable with concentrated buyer spend
Fitness Equipment Revenue Up 55%
Fitness Equipment Brand
Challenge
A high-ticket fitness equipment brand needed to increase revenue while protecting margin. Low-ticket accessories were inflating ROAS but not driving profit.
$90K
Monthly Revenue
+55%
4.5x
ROAS
from 3.6x
$750
AOV
Results
- Revenue increased 40-70% with better AOV mix
- ROAS improved from 3.6 to 4.5
- Profit per order increased without relying on constant sales
Specialty Retailer PMAX ROAS Hits 5.1
Specialty Retailer
Challenge
PMAX was over-indexing on low-intent queries for a specialty retailer with 800-1,500 SKUs. The algorithm was spending freely without profitable returns.
5.1x
ROAS
from 3.0x
$300K
Monthly Revenue
+82%
$58K
Monthly Spend
Results
- ROAS improved from 3.0 to 5.1 within eight weeks
- Revenue increased while wasted spend decreased
- PMAX became a predictable scale lever
Gifting Brand Scales 4x, CPA Holds Steady
DTC Gifts Brand
Challenge
A gifting brand needed to scale through seasonal demand without CPA drifting upward. Ideas and DIY searches were eating budget without converting.
$35K
Monthly Revenue
+60%
$20
CPA
held
4x
Spend Scaled
Results
- Spend scaled 4x while CPA stayed steady at $18-22
- Total sales grew 60%
- Account stayed stable before, during, and after seasonal spikes
Ready to grow with Performance Max for E-commerce?
Get a free audit and see how we can improve your campaign performance.
Make Your Move