Articles · Paid media
Google Ads Performance Max campaign setup: A Practical Workshop
A Performance Max campaign can be configured correctly in the interface and still pursue the wrong business outcome. If a low-intent form submission counts as success, stronger creative and a larger budget do not repair the underlying instruction. They give the campaign more resources to pursue it.
This Google Ads Performance Max campaign setup workshop starts with that instruction, then works through campaign structure, bidding, creative assets, audience signals, landing pages, controls, and launch decisions. Keep a working document beside your Google Ads account: you will use it to record both the settings and the commercial reasons behind them.
Google describes Performance Max as a goal-based campaign type that can access inventory across Search, YouTube, Display, Discover, Gmail, and Maps. It combines advertiser inputs with Google AI to optimise towards conversions or conversion value. That breadth is useful when it serves a clear objective; it is not a substitute for one. explains the underlying mechanics.
Workbench 1: Decide whether this is the right campaign to build
Before opening the campaign builder, answer one question: what should Performance Max contribute that your current acquisition programme does not?
A useful answer might be reaching potential customers beyond existing keyword-based Search campaigns while evaluating them against the same qualified-enquiry definition. For a retailer, it might be promoting an eligible product range against a defensible sales-value objective.
“Google recommends it” is not a campaign brief. Neither is “we need cheaper leads.” Both leave the system’s objective and your evaluation criteria unresolved.
Performance Max deserves consideration when you have a specific conversion goal, suitable creative material, functioning measurement, and a willingness to advertise across Google’s channels. Google explicitly positions it as complementary to keyword-based Search, not simply as a mandatory replacement.
Pause the build if any of these conditions applies:
- Nobody can explain what the selected conversion action represents.
- Sales cannot distinguish a useful enquiry from spam or an unsuitable request.
- The advertised offer is unavailable or inconsistent across landing pages.
- The business requires placement-level certainty that the proposed configuration cannot provide.
- There is no owner for checking expenditure, policy issues, and lead handling after launch.
These are operational readiness criteria, not Google eligibility thresholds. A campaign might technically publish despite them. The issue is whether spending money would generate useful evidence.
Workshop output: write a two-sentence brief naming the commercial outcome, eligible offer, geographic market, and reason for adding Performance Max. Add one sentence explaining what it must not optimise for.
Workbench 2: Write the conversion contract
Treat the conversion definition as a contract between marketing, finance, and the team receiving the demand.
For each proposed conversion action, record what happens, why it matters, where it is measured, and whether it should influence campaign optimisation. Do not assume every action already present in the account belongs in this campaign’s goals.
A page view, resource download, booked consultation, and completed purchase represent different behaviours. Mixing them without a clear rationale makes a total conversion count commercially ambiguous.
For lead generation, separate capture from qualification
Hypothetical example: a SaaS business sells scheduling software to multi-location operators. Its website receives demo requests from suitable businesses, students researching software, and individuals looking for a free personal calendar.
Counting all three as equivalent leads would conceal the acquisition problem. Before launch, the team defines a qualified enquiry as a genuine business request within the supported market and use case. It assigns someone to classify enquiries consistently in the CRM.
Choose the optimisation goal based on the reliable measurement actually available. If only submitted forms are currently measurable, acknowledge that limitation. Establish the qualification workflow before presenting form cost as evidence of customer acquisition efficiency.
Where a downstream conversion integration is planned, make validating it a separate implementation task. Do not select an event merely because its name sounds closer to revenue.
For ecommerce, establish what the value means
Decide whether your working commercial report uses gross order revenue, net revenue after returns, or another clearly defined measure. Then document what the value sent to the advertising platform represents. Those definitions may differ, but the difference must be visible.
Check currency handling, order duplication, and whether cancelled or refunded purchases are accounted for in your evaluation process. A value-based objective is only useful when the values have a consistent meaning.
For implementation detail, use the separate . Here, the task is to approve the campaign’s measurement contract before selecting its goals.
Run a consent-aware test journey
Complete the intended action through the website and confirm that the business receives it correctly. Test relevant consent states and verify that measurement behaviour matches your consent implementation and applicable requirements.
Do not place names, email addresses, phone numbers, or free-text enquiry content in analytics events, campaign parameters, or page URLs. Keep identifying customer information within appropriately governed systems; use non-identifying classifications for aggregate campaign analysis.
Workshop output: an approved conversion definition, a named measurement owner, and evidence that the intended journey works. An unresolved tracking discrepancy is a reason to delay launch, not a footnote for later.
Workbench 3: Set the economics before selecting a target
Google’s documentation explains that Performance Max uses Smart Bidding around your conversion goals and can use an optional CPA or ROAS target. The choice should follow the business objective and measurement quality-not a preference for whichever target looks most ambitious.
Use a conversion-count objective when the selected actions have sufficiently comparable commercial meaning. Consider a conversion-value objective when you have reliable values and meaningful differences between outcomes.
A target is not a guarantee. Entering the cost per acquisition you wish the business could achieve does not establish that the campaign can acquire customers at that cost.
Work backwards from an affordable outcome
Hypothetical illustration, not a benchmark: suppose a business approves an acquisition allowance of ₹12,000 per customer. Its planning assumption is that one in eight qualified enquiries becomes a customer.
The implied allowance is ₹12,000 ÷ 8 = ₹1,500 per qualified enquiry. If only half of raw enquiries qualify, the equivalent raw-enquiry allowance is ₹750.
These numbers are only as dependable as the assumptions. They also exclude other acquisition expenses unless the business has included them in the allowance. Their purpose is to expose the relationship between media cost, qualification, and sales-not prescribe a bid target.
If historical data exists, compare those assumptions with observed outcomes from comparable offers and markets. If it does not, define an affordable exploration budget and label the economics provisional.
Separate daily configuration from financial governance
Record the daily budget you intend to configure, the total expenditure the business is prepared to review, and who can authorise changes. Establish a monitoring cadence and an intervention point rather than treating the daily setting as the entire spending policy.
Google provides budget recommendations and forecasts during setup. Use these as planning inputs, not commitments to revenue, lead volume, or profitability.
Avoid a universal “minimum budget” formula. The useful budget depends on acquisition costs, conversion timing, the scope of the offer, and the evidence needed to make the next decision. If the available amount cannot support that investigation, narrow the campaign brief rather than launch an oversized structure with inadequate funding.
Workbench 4: Build structure around commercial differences
Now enter the campaign builder. Select the relevant objective and Performance Max campaign type, then review the conversion goals presented for this campaign. Interface labels can change; the critical task is confirming that the configuration matches your written contract.
Use a naming convention that identifies the market, offer, and objective. Keep personal information out of campaign names and tracking parameters.
Next, decide what belongs in one campaign and what needs separation. A proposed practical rule is to separate when a difference requires independent budget ownership, substantially different economics, or distinct geographic and operational constraints.
Do not create a separate campaign for every minor product variation. Equally, do not force unrelated offers into one campaign just because they share a website.
Give each asset group a coherent promise
Google defines asset groups as themed collections of text, images, logos, and videos that its system can mix and match into ads. Build them around a customer problem or product theme that can support consistent creative combinations.
Hypothetical example: an office-furniture retailer sells ergonomic chairs and meeting tables. A chair asset group could combine posture-related messaging, chair imagery, relevant video, and a chair-category destination. A table group would use meeting-room messaging and matching products.
A headline about ergonomic seating should not depend on a specific image being present to make sense. Review combinations as if any compatible assets might appear together.
For product-based campaigns, Google specifically recommends aligning product filters with the items on the asset group’s landing page when reviewing generated video alignment. Treat this as a practical consistency check: the product advertised must be findable at the destination.
Workshop output: a campaign-and-asset-group map showing the offer, customer need, creative theme, destination, and reason for each separation. If the reason is only “more granularity,” reconsider it.
Workbench 5: Assemble assets that can survive recombination
Prepare creative before you reach the final publishing screen. Follow the current interface requirements for file specifications and supported formats rather than relying on remembered dimensions or old upload limits.
For each asset group, gather approved text, suitable images, logos, video, and destination URLs. Maintain a simple register showing the owner, approval status, usage rights, and any expiry date for time-sensitive claims.
Build text around distinct functions:
- Offer: what the customer can actually buy or request.
- Use case: the problem or situation the offer addresses.
- Evidence: a verifiable product or service attribute.
- Qualification: a relevant limitation, audience fit, or service condition.
- Action: what happens when someone responds.
Hypothetical example: a software provider might use “Explore Multi-Location Scheduling” as an offer-led message and “Request a Product Walkthrough” as an action. “Cut Admin Time in Half” should not be included without substantiation applicable to the advertised product.
Do not fill every available field with paraphrases of the same claim. Variation should introduce useful information, not just different adjectives.
Review automation as part of creative approval
Google documents text customisation as a campaign-level setting that can generate additional headlines and descriptions from landing-page content, the domain, and existing ads. It also describes auto-generated video assets.
Decide deliberately whether these capabilities fit the business’s approval requirements. A tightly regulated or frequently changing offer may need a more restrictive review approach than a stable, straightforward product range.
For every generated asset you can inspect, check factual accuracy, product availability, brand presentation, and destination consistency. Google makes advertisers responsible for compliance and accuracy; automation does not transfer that responsibility.
An asset report can help identify what to investigate, but it does not replace judgement. Different assets may have different serving opportunities. Do not interpret every performance difference as proof that one creative concept caused more sales.
Workbench 6: Add signals without mistaking them for fences
Google describes audience signals as information about customer intent and preferences that guides learning. Treat them as informed starting points, not a complete definition of everyone who may receive an ad.
Start with what the business genuinely knows: relevant customer needs, product categories, and high-intent behaviours. If considering customer-derived data, first verify the permissions, platform requirements, and legal basis for that use. Possessing a customer list does not automatically make every advertising use appropriate.
Search themes are another way to guide the campaign towards valuable intent. Write them from the buyer’s perspective rather than copying internal product labels.
Hypothetical example: a commercial property consultancy could propose themes around warehouse leasing and industrial space requirements. Before adding them, the team should verify that those services and locations are actually covered by the advertised offer.
Keep a distinction between guidance and restriction. Google identifies search themes as guidance, while negative keywords and brand exclusions are controls for unwanted queries or overlap. Check the current scope of each control before relying on it.
Avoid building an elaborate signal set simply to make the setup look sophisticated. Each input should answer a specific question about who is likely to need the offer and why. Record the reasoning so later reviews can challenge assumptions rather than merely add more signals.
Workbench 7: Control destinations, geography, and brand exposure
Final URL expansion is a consequential setting, not a housekeeping detail. Google says that when enabled, it may replace the supplied final URL with another relevant landing page based on the query and generate matching dynamic text.
Audit the site before enabling that freedom. Review product pages, outdated promotions, recruitment pages, support material, informational articles, and unsupported service areas. Identify which destinations are commercially appropriate and which are not.
If the website contains conflicting offers or stale content, a more restricted destination approach may be preferable while the site is repaired. That sacrifices some flexibility in exchange for greater message control. Verify the available URL controls in the current account rather than assuming that providing one landing page prevents expansion.
For every approved destination, check:
- The advertised offer is visible and available.
- The customer can understand eligibility and next steps.
- Forms or purchasing paths work on mobile.
- Important conditions are not contradicted elsewhere on the page.
- The conversion journey follows the approved measurement and consent design.
Review geographic and language settings against actual service capability. Inspect the location options presented in the interface; do not assume selecting a place name answers every question about geographic reach.
Then review brand exclusions, negative keywords, placement exclusions, and content suitability settings where applicable. Google lists these among Performance Max’s available controls, but their purposes differ. A brand exclusion does not perform the same job as a content suitability filter.
Document the relationship with existing Search campaigns
Google explains that Search campaigns with an exact-match keyword are prioritised when a query matches that keyword, while noting that budget-limited Search campaigns can allow Performance Max to serve on exact-match terms. Search themes have the same priority as phrase- and broad-match keywords.
Consequently, do not assume Performance Max will only reach entirely new demand. Decide how branded demand should be treated and review the combined account picture.
A campaign can report conversions while shifting attribution or demand capture between campaigns. That is different from demonstrating additional business generated by advertising.
At the publish screen: Require a second-person review
Have someone other than the builder compare the saved configuration with the working document. This is a proposed operational safeguard, not a platform requirement.
Ask that reviewer to explain, without prompting, what the campaign is trying to achieve. If they cannot identify the conversion goal, offer, market, and expenditure owner, the handover is incomplete.
Preview available ads and examine asset combinations. Open the landing pages on mobile. Confirm the intended conversion actions, values, budget, bidding approach, destinations, exclusions, and automation settings.
Record the launch configuration and unresolved limitations. A note such as “optimising for submitted enquiries until qualification measurement is validated” is useful because it constrains later claims about success.
Publish only when remaining uncertainties are acceptable business risks rather than preventable defects. Broken forms, inaccurate claims, and misconfigured goals are not experiments worth funding.
After launch: Diagnose before changing
Separate operational checks from performance decisions. Immediately investigate tracking failures, unavailable offers, policy problems, and destination mismatches. Do not wait for a reporting period to end before fixing a demonstrable defect.
Performance assessment needs a different rhythm. Consider the business’s conversion delay and sales cycle before judging recent expenditure. There is no universally valid number of days or conversions that guarantees a dependable decision.
Google lists asset-level metrics, asset-group reporting, channel performance, placement reports, and conversion tracking among Performance Max’s reporting options, depending on campaign and goals. Use these views to answer questions rather than collect screenshots.
A practical review moves through three layers:
- Delivery: is the campaign serving, and are material issues visible?
- Measurement: do reported actions correspond to real business activity?
- Economics: are qualified outcomes and their value acceptable relative to expenditure?
If reported leads increase while sales-accepted enquiries do not, investigate classification, spam, offer wording, and conversion selection before increasing budget. The separate guide to addresses cost diagnosis; this workshop’s priority is ensuring the campaign was built to pursue a meaningful lead in the first place.
For each material change, record the observation, hypothesis, action, expected business effect, and review condition. Avoid changing budget, goals, creative, and destinations simultaneously unless an urgent correction demands it.
Even disciplined before-and-after comparisons are not causal proof. Seasonality, sales follow-up, promotions, and changes elsewhere in the account can influence outcomes. If incrementality matters to a major investment decision, discuss a suitable experimental design and its limitations rather than declaring all attributed conversions incremental.
How Anurag would deliver this implementation
Through his , Anurag would approach Performance Max setup as a connected measurement, creative, and commercial implementation-not just campaign creation.
Inputs: he would request appropriate account access, conversion definitions, existing campaign data, product or service economics, approved assets, landing-page inventories, geographic constraints, and the CRM stages used to evaluate enquiries. Access would be scoped to the work, without collecting unnecessary customer information.
Actions: he would reconcile selected goals with actual business outcomes, inspect the measurement journey, propose campaign and asset-group structure, review budget assumptions, and document audience guidance, URL behaviour, exclusions, and creative automation decisions. Missing prerequisites would become named tasks with owners rather than hidden launch compromises.
Outputs: the engagement would produce a campaign configuration record, conversion contract, asset-and-destination map, unresolved-risk register, and reporting plan. Launch approval would show what is being optimised, what remains provisional, and who owns the next decision.
Measurement: review would pair platform reporting with available sales or order-quality evidence. The emphasis would be on qualified acquisition cost, reliable conversion value, expenditure control, and operational issues-not a promised performance uplift.
The value of that process is a clearer relationship between business intent and the instructions given to automation. Results still depend on demand, competition, offer quality, measurement, and execution beyond the campaign itself.
Leave the workshop with a decision record
Your finished setup should be explainable in plain language: “We are promoting this offer, to this serviceable market, towards this measured outcome, within this spending policy, using these approved assets and destinations.”
If any part remains vague, return to that workbench before adding complexity. If the configuration is ready, agree on the first review questions and who will answer them. Launch is the start of evidence collection, not proof that the commercial assumptions were correct.
For help turning those decisions into an account-specific implementation, with your business objective, current measurement limitations, and intended offer. Those inputs are more useful than asking whether Performance Max is universally better than Search.
Sources
- - campaign inventory, conversion goals, bidding inputs, asset groups, signals, search prioritisation, URL expansion, creative automation, controls, and reporting options.