Performance Max reports a strong return on ad spend. Yet total orders, new customers or qualified opportunities have barely increased.
Before raising the budget, ask what the campaign is contributing beyond the demand the business already had. Advertising attribution tells you which outcomes a platform credits to its ads. It does not directly establish what would have happened without them.
This review is about additional business outcomes. It requires a different question from whether the reported revenue covers advertising cost.
Keep three questions separate
Branded search, returning customers and incremental sales are related concepts, but they are not interchangeable.
| Question | What it describes | What it cannot prove alone |
|---|---|---|
| Was the query branded? | The person searched for a particular business or product brand | Whether the customer has bought before or would buy without the ad |
| Was the customer new? | The outcome meets a documented new-customer definition | Whether advertising caused that first purchase |
| Was the sale incremental? | The sale would not have happened under the relevant comparison condition | Which channel deserves all the credit for the customer journey |
A new customer can search for your brand after a recommendation. An existing customer can buy a product they would not otherwise have considered because of an advertisement.
Avoid labelling every branded conversion unnecessary, or every first purchase proof of advertising growth.
Put the campaign report beside the business ledger
Choose comparable periods and write down the reporting basis. Check time zones, attribution settings, conversion delays and whether sales are recorded on an interaction or transaction date.
Compare the campaign with total orders or opportunities, distinct new customers and relevant sales outcomes in the business records. Use consistent treatment of cancellations and returns.
The following is a hypothetical reporting pattern, not evidence about a real account:
| Measure | Period A | Period B |
|---|---|---|
| Performance Max spend | $2,000 | $3,000 |
| Revenue attributed to Performance Max | $12,000 | $21,000 |
| Total store orders | 400 | 405 |
| New purchasing customers in store records | 120 | 121 |
The campaign report improved materially while the business counts barely changed. That pattern deserves investigation. It does not establish causation: order value, seasonality, other channels, product availability and reporting changes could also explain the difference.
The next step is to identify plausible explanations and design a comparison that can distinguish them.
Review branded demand with the evidence available
Inspect the search reporting available for the campaign and the business's branded Search activity. Record coverage limitations rather than assuming that the visible queries describe all traffic.
Look for changes in branded demand, overlap with other advertising and the destinations used. A creator campaign, email promotion or offline recommendation may increase people searching for the business.
Google provides brand exclusions for Performance Max, with scope limited to specified inventory and an option affecting Shopping treatment.1 Review the current settings before applying an exclusion. It is a targeting control, not an automatic incrementality test.
If you exclude branded demand from one campaign and another paid campaign captures it, the first campaign's ROAS may fall while total spending still purchases similar demand.
Validate new-customer reporting
Define “new customer” using the business's own records and a clear observation period. Account for offline purchases, multiple sales systems and duplicate customer records where relevant.
Then review how the advertising system receives or infers customer status. An incomplete customer list or missing purchase history limits the usefulness of the classification.
Google's customer lifecycle goals include modes that prioritise new customers or focus bidding on them. Requirements vary by mode, bidding strategy and conversion configuration.2 Confirm eligibility and data readiness before changing the campaign.
The setting can align bidding with an acquisition objective. It does not create a randomized comparison showing how many first purchases would have occurred without advertising.
Write a test around one business question
A useful hypothesis is specific:
“We believe the current campaign captures substantial existing demand. We want to test whether a defined change improves additional new-customer outcomes at an acceptable cost.”
Choose the measurement design according to scale and practical constraints.
| Approach | Useful question | Main limitation to record |
|---|---|---|
| A suitable holdout or lift study | What outcomes occur with and without the tested advertising exposure? | Eligibility, sample size, spillover and study design matter |
| Comparable geographic test | Do business outcomes differ between appropriately selected test and comparison areas? | Regions may differ, and customers or other media can cross boundaries |
| Campaign targeting experiment | Does the tested targeting configuration produce a better customer mix or economics? | It may compare configurations rather than advertising versus no advertising |
| Before-and-after observation | What changed following a documented intervention? | Seasonality, other campaigns and external changes can confound the result |
A low-volume business may not have enough data for a reliable lift study. In that case, use the strongest feasible evidence and label the conclusion appropriately.
Changing brand exclusions, budget, bidding and customer acquisition settings together makes the result harder to interpret.
Measure the whole business response
Before launch, define the primary outcome, spending boundary, review conditions and guardrails.
The primary outcome might be additional new purchasing customers or qualified new opportunities under the selected design. Also inspect contribution after acquisition cost, overall sales and whether another paid channel absorbed the demand.
Do not choose the winner solely because its platform ROAS is higher. A campaign can lose easy attributed conversions while the business becomes more effective at acquiring additional customers.
Equally, a lower ROAS is not evidence of greater incrementality. It still needs business outcomes that justify the spending.
Decide what job Performance Max should do
The result may support an acquisition focus, a deliberate retention role, a change in branded traffic handling or a smaller investment while better evidence is gathered.
Write the objective plainly and make the reporting match it. Profitable returning-customer sales can be valuable when retention is the intended job. They should be labelled accurately when the business is evaluating new-customer growth.
A Google Ads review can connect campaign controls to that objective. If the numbers do not reconcile, use the CRM and ad-report guide first. If the separate concern is margin, review high ROAS and low profit alongside this growth analysis.