Your Authoritative Guide to Google Performance Max (PMax) Campaigns

We want to test PMax. What's the minimum budget required to see results?

There is no official "minimum spend" set by Google, making PMax accessible for various budgets. However, the key is to provide the algorithm with enough data to exit its learning phase and optimize effectively. A budget that is too low can lead to a slow ramp-up and fewer conversions.

Our strategic approach focuses on goals rather than a fixed number:

• For initial testing: A daily budget of at least $50-$100 can be a reasonable starting point to gather sufficient data.
• Based on your goals: A common best practice is to set a daily budget that is at least three to ten times your target cost-per-acquisition (CPA). This ensures the campaign has enough room to learn and find conversions.
• For scaling: Once a campaign proves profitable—for instance, by consistently achieving a target Return on Ad Spend (ROAS) of over 500%—we can confidently recommend increasing the budget to capture more opportunities. New campaigns are inherently risky and often require more investment upfront to generate the necessary performance data.

Ultimately, the goal is to fund the campaign adequately so it can move through its learning phase and stabilize, which is when you'll begin to see consistent results.

What are the essential assets we need to provide for a PMax campaign?

Providing a rich and diverse set of high-quality creative assets is one of the most critical factors for PMax success. While Google can auto-generate some assets if they aren't provided, this is not optimal. Supplying your own creative ensures brand consistency and significantly better performance.

Here are the essential assets required for a strong PMax campaign:

• Text Assets:
• Headlines: Up to 15 headlines of 30 characters each. Provide variety and include at least one with 15 characters or less.
• Long Headlines: Up to 5 long headlines of 90 characters each.
• Descriptions: Up to 5 descriptions. Four should be 90 characters, and one should be 60 characters or less.
• Business Name: Your company name, up to 25 characters.
• Image Assets: A maximum of 20 images are allowed. It is highly recommended to provide a variety of orientations to ensure your ads can serve across all of Google's inventory.
• Landscape (1.91:1): At least one, with a recommended size of 1200x628.
• Square (1:1): At least one, with a recommended size of 1200x1200.
• Portrait (4:5): Recommended, with a size of 960x1200.
• Logos: You can add up to 5 logos.
• Square (1:1): At least one is required, with a recommended size of 1200x1200.
• Landscape (4:1): Optional, but recommended, with a size of 1200x300.
• Video Assets: While optional, videos are "really important" for asset group strength and are highly recommended. If you don't provide a video, Google will create a low-quality auto-generated one from your other assets.
• Provide at least one video that is longer than 10 seconds.
• For best results, include multiple videos of different lengths (e.g., a short bumper under 6 seconds, a 10-15 second ad, and a 30+ second ad) and orientations (horizontal, vertical, square).

How does PMax use audience signals, and can we integrate our 6sense data?

Audience signals are a crucial way to guide PMax's machine learning, even though they are not used for direct, restrictive targeting. Think of them as "suggestions" or a "nudge" that helps the algorithm understand your ideal customer profile, which speeds up the learning phase and improves efficiency. PMax will still show ads to users outside your signals if it identifies them as likely to convert.

There are several types of audience signals you can use:

• Your Data (First-Party Data): This is the most powerful signal you can provide. It includes remarketing lists of website visitors, app users, and customer match lists (e.g., email lists of prospects or existing customers).
• Custom Segments: You can create audiences based on search activity, visited websites, or app usage. For example, you can target people who search for specific competitor keywords or browse industry-related sites.
• Interests & Demographics: These are Google's predefined audiences, including in-market segments (users actively researching products) and affinity audiences (users with specific interests and habits).

Integrating 6sense Data: Yes, you can and should integrate data from platforms like 6sense. This is typically done by exporting target account lists or lists of users who have shown high intent and uploading them to Google Ads as a customer match list. This list then becomes a powerful first-party data audience signal, telling PMax to look for users who share characteristics with your highest-value prospects.

We've heard PMax can lead to a lot of invalid traffic. How do we mitigate this?

Invalid traffic (IVT) and low-quality leads are valid concerns with PMax due to its automated nature and broad reach. We have seen instances where PMax campaigns generated "spammy" or "junk" leads, which ultimately led to pausing the campaigns in favor of more controllable formats like Demand Gen.

However, there are several strategies to mitigate this risk:

• Use Strong Audience Signals: Providing high-quality first-party data (like customer lists) helps the algorithm focus on more relevant users from the start.
• Implement Form Validation: Ensure your website forms have robust validation, such as double validation or CAPTCHA, to filter out bots before they are even recorded as a lead.
• Account-Level Negative Keywords: You can add negative keywords at the account level to prevent your ads from showing on irrelevant or low-intent search queries.
• Placement Exclusions: While you can't control placements directly within PMax, you can monitor placement reports to see where your ads are appearing. If you find low-quality websites or apps, you can exclude them at the account level.
• Content Exclusions: Use brand suitability controls to exclude sensitive content categories, preventing your ads from appearing alongside content that doesn't align with your brand.
• Third-Party Tools: For an added layer of protection, specialized ad fraud prevention tools can detect and filter invalid traffic in real-time, feeding exclusion lists back to Google to refine the algorithm's targeting.

How long is the 'learning phase' for a PMax campaign, and when can we expect to see stable performance?

The learning phase is the period where Google's algorithm gathers data to learn how to best achieve your campaign goals. During this time, performance can be volatile.

Based on our experience and Google's guidance, the learning phase for a PMax campaign typically lasts between four to six weeks. One of our internal projects noted that a recreated asset group needed 14 to 30 days to stabilize again.

It is critical not to make significant changes to the campaign during this initial learning period, as this can reset the algorithm and prolong the phase. After about six weeks, you should have enough data to make an initial assessment of the campaign's performance and begin making strategic optimizations.

Can we use PMax to target specific industry verticals or just broad audiences?

Yes, you can use PMax to target specific industry verticals, but it requires a strategic approach. PMax was not originally designed with niche B2B targeting in mind, so you must provide strong signals to guide its broad automation.

Here are the best methods for targeting verticals:

• Use Audience Signals: The most effective way is to use your first-party data. Create customer lists of clients within a specific vertical and use them as an audience signal. You can also build custom segments based on users who visit industry-specific websites or search for industry-related keywords.
• Thematic Asset Groups: Structure your asset groups around specific industries. This allows you to tailor your ad copy, images, and videos to resonate with that vertical's unique pain points and language.
• Start Broad, Then Segment: A common challenge is that overly narrow audiences can cause the algorithm to underperform or struggle to gather enough data. A sound strategy is to start with a slightly broader audience and then, based on performance data, create new, more segmented asset groups for the verticals that are performing best.

How much control do we have over where our PMax ads are shown?

PMax is Google's most automated campaign type, which means advertisers have the least direct control over ad placements compared to traditional campaigns. Your ads can appear across Google's entire inventory, including Search, Shopping, YouTube, Display, Gmail, Discover, and Maps.

While you cannot choose specific channels (e.g., "only run on YouTube"), you can influence and restrict placements through several methods:

• Asset-Driven Influence: The assets you provide guide the algorithm. For example, providing high-quality videos makes it more likely your ads will serve on YouTube, while a strong product feed prioritizes Shopping.
• URL Expansion Setting: Turning off "Final URL Expansion" in the campaign settings prevents PMax from sending traffic to pages on your site other than the specific landing page you've designated. This gives you tighter control over the user's destination.
• Account-Level Exclusions: This is your primary tool for control. You can create exclusion lists at the account level for specific websites, YouTube channels, or mobile apps where you don't want your ads to appear. PMax will respect these account-level exclusions.
• Brand Exclusions: You can prevent your ads from showing for specific brand-related search queries.

What are the best practices for structuring asset groups within a PMax campaign?

A well-organized asset group structure is fundamental to testing, optimizing, and scaling PMax campaigns. Instead of putting all assets into a single group, we recommend a more segmented approach.

Here are the best practices we follow:

• Structure by Theme or Category: The most effective strategy is to create separate asset groups for different product categories, service lines, or customer personas. For example, one asset group could be for "melatonin supplements" and another for "multivitamins." This allows you to tailor all assets—headlines, images, and videos—to a specific theme, which improves relevance and performance.
• Create Separate Groups for A/B Testing: When you want to test new creatives or messaging, create a new asset group. This allows you to A/B test against your control group without disrupting the performance of an already successful asset group.
• Ensure Sufficient Conversion Volume: Each asset group needs enough data to optimize effectively. A good rule of thumb is to aim for at least 20-30 conversions per asset group per month. If an asset group is struggling to get conversions, consider merging it with another to consolidate data.
• Use All Asset Types: For each asset group, provide a full suite of assets (text, images in all orientations, and videos). This gives the algorithm maximum flexibility to create and test different ad combinations across all channels.

How do we measure the success of a PMax campaign compared to traditional search or display?

Measuring PMax success requires a more holistic view than single-channel campaigns because it operates across the entire marketing funnel. While a traditional search campaign might be judged solely on direct, last-click conversions, PMax's impact is often broader.

Here are the key metrics and considerations for measuring PMax success:

• Primary Conversion Metrics: Like any performance-focused campaign, the core KPIs are Cost per Acquisition (CPA), Return on Ad Spend (ROAS), and total Conversion Value or Revenue. These metrics tell you if the campaign is efficiently driving your main business goals.
• Assisted Conversions: PMax often acts as an introductory touchpoint. A user might see a YouTube ad via PMax and later convert through a branded search. It's crucial to analyze assisted conversions to understand the campaign's full impact on the customer journey.
• Lead Quality vs. Quantity: For B2B or lead generation, success isn't just about the number of form fills. It's essential to track the quality of those leads. We analyze the progression from lead to Marketing Qualified Lead (MQL), to Sales Qualified Lead (SQL), and ultimately to closed-won opportunities and pipeline revenue. A high volume of "junk leads" indicates a problem, even if the top-line CPA looks good.
• Impression Share: This metric helps you understand your visibility and if there's room to scale by increasing your budget.
• Uplift Tests: To truly isolate PMax's incremental value, you can run uplift experiments. This involves comparing a group that sees only your standard search campaigns to a group that sees both search and PMax, allowing you to measure the true "lift" in conversions.

Can we exclude our brand keywords from PMax campaigns to avoid cannibalization?

Yes, you can and should control how PMax handles your brand keywords. Google provides a brand exclusion feature at the campaign level that prevents your PMax ads from serving on search queries for your specified brand terms.

This is a critical strategic decision:

• Argument for Exclusion: Many advertisers choose to exclude brand terms to avoid paying for clicks from users who are already searching for them and would likely convert organically. This is often seen as a way to prevent budget waste.
• Argument for Inclusion (Brand Defense): If competitors are actively bidding on your brand terms, not running your own branded ads can mean you lose the top ad spot to a competitor. In this scenario, running branded ads—either within PMax or in a separate search campaign—becomes a necessary defensive strategy to protect your traffic.

Our approach is to have a direct discussion about this. If brand protection is a priority due to aggressive competition, we may recommend including brand terms. Otherwise, we can apply exclusions to focus the PMax budget on non-branded discovery and acquisition.