Performance Max can spend a significant budget, generate sales, and report a positive ROAS while still failing to create meaningful profit for an eCommerce business.
Performance Max does not optimise directly for profit. It optimises for the conversion goals and values configured inside Google Ads. If those inputs do not reflect the real economics of the business, the campaign can perform exactly as instructed while producing the wrong commercial outcome.
This is why a campaign should not be judged only by whether it spends, generates conversions, or reaches its target ROAS.
What Performance Max actually optimises
Performance Max is a goal-based campaign type that can deliver ads across Google Search, Shopping where available, YouTube, Display, Discover, Gmail, Maps, and other eligible Google inventory. It uses Google’s automated bidding and delivery systems to decide when, where, and to whom ads should appear.
This gives the campaign access to far more signals and placements than a manager could control manually in real time.
However, Google still needs a goal.
For a typical eCommerce campaign, that goal should usually be a completed purchase. The system then looks for opportunities that appear likely to generate more purchases or more conversion value within the available budget and bidding constraints.
The problem begins when the selected goal does not represent the result the business actually needs.
If Add to Cart, Begin Checkout, page views, contact actions, and purchases are all included as primary conversion goals, Google may optimise towards a mixture of actions with very different commercial value.
An add-to-cart action is easier to generate than a completed order. If both are treated as important bidding outcomes, the campaign may find more users who add products to their baskets without becoming paying customers.
For a standard eCommerce business, Purchase should usually be the primary optimisation goal. Micro-conversions can remain available as secondary actions for analysis, but they should not normally carry the same bidding importance as revenue-generating orders. Google Ads allows advertisers to choose which conversion goals are used for bidding at the campaign level.
There can be exceptions, particularly when a business has very low purchase volume or an unusual buying process. They should remain exceptions rather than the default setup.
Performance Max cannot distinguish a commercially valuable action from an unimportant one unless the account tells it which result matters.
Revenue, ROAS, and profit are not the same result
Performance Max typically works with conversion value, rather than the business’s complete profit calculation.
If an order worth EUR 200 is sent to Google Ads, the platform sees EUR 200 in conversion value. It does not automatically subtract the product cost, payment fees, fulfilment, shipping subsidies, discounts, returns, taxes, or other expenses attached to that order.
This means a strong reported ROAS may still sit below the point at which the business makes money.
The gap becomes even more important when the store sells products with different margins.
A campaign may generate the same revenue from two products while creating very different financial outcomes. One product may have a healthy margin and a low return rate. Another may have expensive fulfilment, frequent returns, and little profit after promotional discounts.
Google will not necessarily prefer the more profitable product. It will prefer the opportunities that best match the conversion value and bidding objective it receives.
This is why break-even ROAS matters.
Break-even ROAS is the minimum return required for advertising revenue to cover the costs included in the business calculation. The exact number depends on the store’s margin structure, fulfilment model, payment costs, return rate, and other commercial factors.
It does not need to become an overly complicated financial model before a campaign can run. But the business should know whether its target ROAS represents actual profitability or simply an attractive number inside Google Ads.
A campaign reporting a 400 percent ROAS is not automatically better than one reporting 300 percent. The first may be selling low-margin products, while the second is generating more contribution profit.
Platform efficiency should always be interpreted through business economics.
Inaccurate conversion data gives Google the wrong target
Performance Max depends heavily on conversion tracking because automated bidding uses that information to understand which auctions and users are producing results.
The campaign does not know whether a Purchase event is correct simply because it received one.
Common problems include:
Purchase events firing more than once
Failed or cancelled orders being recorded as completed purchases
Incorrect order values or currencies
Test transactions entering production data
Several conversion actions measuring the same sale
Micro-conversions being included as primary bidding goals
Large differences between Google Ads, GA4, and actual store orders
Duplicated purchases can make the campaign appear far more profitable than it is. If one real order is recorded twice, Google sees more revenue, a lower cost per purchase, and a stronger ROAS.
The system may then spend more aggressively based on results that never happened.
Google Ads, GA4, and the store backend will not always show identical numbers. They use different attribution rules, event collection methods, and reporting logic. Small, explainable differences are normal.
Large or sudden discrepancies are not something to ignore.
Tracking does not need to be perfect before every optimisation decision, but it must be reliable enough to represent the direction of the business. Otherwise, every bidding decision is based on a distorted version of performance.
A separate conversion tracking strategy is required to cover implementation, consent, Enhanced Conversions, attribution, and technical validation in depth. For Performance Max, the essential point is simpler:
The campaign can only optimise towards the sales data it receives.
Target ROAS does not create profitability
Target ROAS is often treated as a profit control.
It is not.
Target ROAS tells Google to seek as much conversion value as possible while aiming for the selected return. Google notes that individual conversions may still produce a higher or lower return than the target.
The target controls the relationship between advertising spend and the conversion value available inside Google Ads. It does not confirm whether that value produces profit after business costs.
Setting an aggressive target can also restrict delivery.
If the target is much higher than the campaign can realistically achieve, Google may enter fewer auctions, reduce spend, and lose conversion volume. The campaign may protect its reported efficiency while limiting growth.
A target that is too low can create the opposite problem. Google receives more freedom to spend, but the additional revenue may not meet the business’s profitability requirements.
There is also a difference between maximising conversion value with and without a target.
Without a target ROAS, Google generally aims to use the available budget to generate as much conversion value as possible. With a target, it tries to balance conversion value with the requested return.
Starting without a strict target can sometimes help a new campaign collect useful data, especially when the account has little reliable conversion history. It is not always appropriate. A business with narrow margins may need stronger controls from the beginning.
There is no bidding setting that removes the need for judgment.
The target should be based on recent, reliable data and the economics of the products being advertised. It should also be reviewed when margins, pricing, promotions, average order value, or market conditions change.
Changing it repeatedly after a few weak days usually creates more noise than clarity.
Brand traffic can make Performance Max look stronger than it is
Performance Max can serve on branded searches unless the campaign is configured to exclude them.
This means it may reach people who are already searching for the company, website, or products by name. These users often convert at a high rate because they already know the brand and may already intend to purchase.
Branded conversions are not bad conversions.
The problem is interpretation.
If a large percentage of Performance Max revenue comes from existing brand demand, the reported ROAS may say more about brand familiarity than the campaign’s ability to attract new customers.
This can also blur the role of a separate brand Search campaign.
Revenue that previously appeared under brand Search may begin appearing under Performance Max. Total business sales may remain similar even though Performance Max now appears to be producing exceptional growth.
Google provides campaign-level brand exclusions for advertisers who want to prevent Performance Max from serving on specific branded searches. Google specifically recommends this control when brand traffic is managed separately and the advertiser wants to avoid paying for the same queries through Performance Max.
Brand exclusions should not be applied automatically to every account.
In some cases, allowing branded traffic inside Performance Max can support coverage and sales. In others, excluding it creates a clearer picture of non-brand acquisition and gives the advertiser more direct control through a dedicated Search campaign.
Brand Search and Performance Max can work together; they simply need defined roles.
A strong ROAS built largely on customers already looking for the brand should not be presented as proof that Performance Max created the same amount of incremental demand.
Existing customers can improve ROAS without improving acquisition
Performance Max may also generate revenue from customers who have purchased before.
Returning customers often convert more easily than new customers. They already trust the store, understand the product, and may need less persuasion before placing another order.
This can improve reported ROAS.
It does not necessarily help a business whose primary goal is customer acquisition.
Google Ads includes customer lifecycle goals that can be used to prioritise new customers, bid more for them, or focus only on new customers in eligible campaign setups. These features use Smart Bidding and customer data to influence optimisation.
The quality of that distinction depends partly on the information available to Google.
Customer Match lists, purchase data, tagging, and Google’s own detection can all contribute to classifying customers. No setup should be assumed to identify every customer perfectly.
For this reason, new customer performance should also be compared with backend, CRM, or eCommerce platform data where possible.
A campaign can have a strong total ROAS and weak new customer economics at the same time.
That does not make existing-customer revenue irrelevant. It means acquisition and retention should not be evaluated as though they are the same objective.
Product priorities need to exist outside the algorithm
Performance Max may naturally direct more spend towards the products it can sell most easily.
Those may not be the products the business most wants to grow.
A product can attract spend because it has a low price, high conversion rate, recognisable name, strong existing demand, or a large amount of historical data. Another product may be strategically more important but require more education or a longer decision process.
Without clear structure, the easier product can absorb the campaign budget.
This is why product groups often need to be separated according to genuine business differences, such as:
Product category
Margin
Price range
Seasonality
Stock availability
Market
Promotional priority
Customer type
Performance history
The separation can happen through different campaigns, asset groups, listing groups where available, landing page controls, or budget decisions. The correct method depends on the market and account setup.
For advertisers in countries where Merchant Center product feeds and Shopping inventory are available, product data becomes one of the most important Performance Max inputs. Titles, identifiers, pricing, availability, images, and product groupings all affect how Google understands and serves the catalogue.
Where feed-based inventory is not available, the underlying principle still applies. Products should not be placed into one automated system without a clear way to evaluate and control their business role.
Automation needs structure around the products, not only around the campaign.
Creative assets influence where Performance Max can compete
Performance Max is not only a Search campaign with additional placements.
It can use creative assets across YouTube, Display, Discover, Gmail, Search, and other eligible inventory. Google describes an asset group as a collection of creative assets built around a theme or target audience.
This means the quality and organisation of the creative matter.
An asset group should have a clear reason to exist. It may be organised around a product group, category, use case, audience need, or campaign message.
Placing generic headlines, unrelated images, several products, and broad landing pages into one asset group gives Google many possible combinations but little strategic direction.
More assets are not automatically better when they all communicate the same vague message.
Useful creative gives Google different ways to present the product across placements and stages of awareness. This may include:
Clear product benefits
Category-specific messages
Strong commercial offers
Product demonstrations
Use cases
Objection handling
Social proof
High-quality images and video
Google recommends supplying a diverse range of relevant assets and provides reporting to help advertisers evaluate their contribution.
Automatically generated video can help a campaign access video inventory, but it is rarely a complete creative strategy for an established eCommerce brand.
The objective is not to fill every available field. It is to give Performance Max enough strong material to represent the products accurately wherever the campaign appears.
The landing page still decides whether traffic converts
Performance Max can identify traffic opportunities.
It cannot make a weak product page convert.
If the landing page is slow, unclear, poorly structured, or inconsistent with the ad, the campaign may generate visits without profitable sales.
The page still needs to answer the customer’s main questions:
What is the product?
Why is it relevant?
What makes it different?
What does it cost?
Is it available?
Can the customer trust the store?
How difficult is delivery and checkout?
Final URL expansion can allow Performance Max to select landing pages beyond the URL initially provided. Advertisers can use URL inclusions and exclusions to guide which parts of the website are eligible. When Final URL expansion is disabled, traffic is restricted to the URLs provided through asset groups and page feeds.
This can increase reach when the website is well organised.
It can also send traffic to pages that are technically relevant but commercially weak, such as outdated categories, unavailable products, informational content, or pages with poor conversion intent.
Automation should therefore be supported by a website structure that clearly separates valuable commercial pages from pages that should not receive paid traffic.
The campaign, ad message, product, and landing page should continue the same argument.
When the ad promises one thing and the page presents another, the cost of that mismatch appears in the conversion rate.
Search themes and audience signals provide direction, not control
Search themes allow advertisers to tell Performance Max which queries they know their customers may use. Google describes them as optional and additive to the traffic the campaign can already identify through assets, landing pages, feeds, and other signals.
They are not traditional Search keywords.
Adding a search theme does not mean Performance Max will target only that query. It gives Google additional context about relevant demand.
Audience signals work in a similar way. They are suggestions that can help guide the campaign towards useful customer patterns, but they are not permanent audience restrictions.
First-party customer lists, relevant custom segments, website audiences, and other high-quality signals can help a campaign understand where to begin.
They cannot repair a poor commercial setup.
A campaign with wrong conversion goals, weak creative, an unclear offer, and poor landing pages will not become profitable simply because more audience signals or search themes were added.
These inputs can point Google in a direction.
They do not define whether the destination is profitable.
Performance Max is more transparent, but the total result still needs context
One historic criticism of Performance Max was limited visibility into where spend and conversions occurred.
Google has expanded channel performance reporting, allowing advertisers to see how a campaign is delivering across Google’s channels and inventory, with additional format-level data and diagnostics.
This is useful because one total ROAS can hide very different types of delivery.
A campaign may rely heavily on Search and Shopping inventory where available. Another may use more YouTube, Display, Discover, or other placements.
Those channels do not always play the same role in the customer journey.
The channel report should therefore be used to understand how Performance Max is finding conversions, not as an instruction to judge every channel only by its final-click sales.
It also does not turn Performance Max into a collection of manually controlled channel campaigns.
Performance Max remains an automated cross-channel campaign. The purpose of greater transparency is to improve diagnosis and decision-making, not to recreate separate Search, YouTube, and Display management inside one campaign.
Performance Max should not be the whole Google Ads strategy
Google describes Performance Max as a campaign type designed to complement keyword-based Search campaigns.
That distinction matters.
Search campaigns give advertisers direct control over keywords, queries, ad groups, copy, and landing pages. They are often useful when the business needs to capture specific demand, protect brand searches, prioritise categories, or manage queries more precisely.
Performance Max provides broader cross-channel reach and automated delivery.
The two campaign types can work together when their roles are clear.
A complete eCommerce Google Ads structure may use different campaigns for:
Branded Search
High-intent non-brand Search
Performance Max
Shopping where available
Demand generation
Remarketing
Specific product or market priorities
Not every account needs every campaign type.
The point is that Performance Max should have a defined role within the wider strategy. It should not become the entire strategy simply because it can access most Google inventory.
Profitable Performance Max requires human direction
Performance Max is powerful because it can process a scale of signals, auctions, placements, and customer journeys that no person could manage manually.
It is risky for the same reason.
The system can move the budget quickly towards the result it has been told to value. If that result is incomplete, inaccurate, or commercially weak, automation can scale the problem.
Strong management, therefore, involves more than changing target ROAS.
It requires connecting:
Conversion goals
Revenue accuracy
Product priorities
Margins and break-even targets
New and returning customer value
Brand demand
Bidding and budgets
Creative assets
Search and audience signals
URL controls
Landing page performance
Wider account structure
At Creative Brackets, we treat Performance Max as one part of the complete eCommerce system. Google Ads cannot be separated from tracking, product economics, website performance, creative, and customer behaviour.
The purpose is not to make Performance Max spend more.
It is to make sure the campaign is learning from the right information and scaling results the business actually wants.
Performance Max does not define profitable growth
Performance Max can find auctions, customers, placements, and conversion opportunities.
It cannot decide what good growth means for the business.
It does not automatically know which products have the strongest margins, which orders are likely to be returned, which customers have the highest long-term value, or whether branded sales represent new demand.
Those definitions must come from the people managing the account and the business behind it.
When a campaign spends without generating profit, the answer is rarely one hidden setting.
The real issue is usually the relationship between the campaign goal, the data Google receives, the products being prioritised, the customers being acquired, and the economics used to judge success.
Performance Max needs automation to operate at scale. It still needs human control to define what is worth scaling.
Creative Brackets manages Google Ads as a connected growth system, combining campaign strategy, conversion quality, bidding, product priorities, creative, CRO, and the post-click experience.
Because a campaign that spends is not necessarily working.
A campaign is working when the growth it creates makes commercial sense.