Performance Max (PMax) has become one of the most discussed — and debated — campaign types in Google Ads. Launched broadly in 2021 and made the default for Smart Shopping campaigns in 2022, it now powers a significant share of total Google Ads spend.
For advertisers who understand how it works, PMax can deliver excellent results. For those who treat it as a set-and-forget solution, it can quietly waste budget. This guide explains what Performance Max actually does and how to run it effectively.
What Is Performance Max?
Performance Max is a goal-based campaign type that uses Google's machine learning to serve ads across all Google inventory from a single campaign: Search, Shopping, Display, YouTube, Gmail, Maps, and Discover.
Rather than setting individual keyword bids or targeting specific placements, advertisers provide:
- Asset groups — combinations of creative assets (headlines, descriptions, images, videos, logos)
- Audience signals — data to guide the algorithm toward likely converters
- A conversion goal — what you want the campaign to optimise for
Google's automation then decides where, when, and to whom to show your ads to maximise conversions within your budget.
The appeal is consolidation: one campaign across all inventory, managed by an algorithm with access to more data signals than any human could process. The risk is opacity: you give up significant control, and the algorithm doesn't always make the decisions you'd make.
How Performance Max Works
PMax uses a combination of signals to determine where to allocate budget:
Audience signals — Inputs from the advertiser (customer lists, website visitor lists, custom intent audiences, similar audiences) that tell the algorithm which user types are most likely to convert. The algorithm uses these as starting points, not hard limits — it will expand beyond your audience signals if it finds conversion opportunities elsewhere.
Creative assets — PMax tests combinations of the assets you provide to determine which combinations perform best in different contexts. Asset quality and quantity significantly affects performance.
Conversion data — The more conversion data the algorithm has to learn from, the better it performs. New accounts or campaigns with fewer than 30–50 conversions per month will underperform accounts with strong conversion histories.
Final URL expansion — By default, PMax can dynamically match searchers to your most relevant landing pages based on their query. This is powerful but requires careful management.
Setting Up Performance Max Effectively
Define conversion goals correctly
PMax optimises for the conversion goals you set. If you include soft conversions (e.g., page views, time on site) alongside hard conversions (purchases, form fills), the algorithm may optimise for the wrong outcomes. Use only your most valuable conversion actions as primary goals.
Build strong asset groups
PMax performs best when it has sufficient creative variety to test across different contexts:
- Minimum: 5 headlines, 5 descriptions, 3–5 images (landscape and square), 1 logo
- Recommended: 15 headlines, 4 descriptions, 7–10 images, 1–3 videos
Video is particularly important. Without video, PMax will auto-generate videos from your images — often with poor results. Even a simple 30-second product or brand video significantly improves YouTube and Display performance.
Use audience signals strategically
Provide rich audience signals that represent your best customers:
- Customer email lists (past purchasers, high-LTV customers)
- Website visitor lists segmented by behaviour (checkout abandoners, product viewers)
- Custom intent audiences based on relevant search terms
- In-market audiences relevant to your product category
The algorithm will use these as learning signals and then expand. Don't be too restrictive — the expansion is often where PMax finds incremental volume.
Brand exclusions
One of the most important PMax configuration steps is excluding your own brand terms. Without brand exclusions, PMax will often spend budget on branded searches that would have converted anyway — inflating apparent ROAS while cannibalising budget from genuinely incremental campaigns.
Apply brand exclusion lists at the account level.
URL expansion controls
Final URL expansion can cause PMax to serve ads for queries and landing pages you haven't vetted. Review the URL expansion report regularly. Use URL expansion exclusions to prevent traffic being sent to pages that aren't appropriate for paid acquisition (blog posts, legal pages, etc.).
Performance Max vs Standard Shopping
For ecommerce, PMax largely replaced Smart Shopping campaigns. Compared to standard Shopping campaigns:
PMax advantages: Access to more inventory (Search, Display, YouTube alongside Shopping), more sophisticated audience signals, and potentially higher volume at scale.
PMax disadvantages: Less transparency and control over product-level bidding, more difficult to diagnose performance issues by product category, and limited negative keyword functionality.
For ecommerce advertisers, a hybrid approach often works well: PMax for volume across all inventory, supplemented by standard Shopping campaigns for tightly controlled brand and top-product terms where you want more manual control.
Reporting and Optimisation
PMax provides limited transparency compared to traditional campaign types. Key available reports:
Asset performance — Which creative assets are rated Low, Good, or Best. Replace Low-rated assets regularly.
Audience insights — Which audience segments are converting, allowing you to build more granular audience signals for future campaigns.
Search terms insight — An aggregated view of search query categories driving conversions (not individual search terms). Limited but useful for identifying irrelevant traffic patterns.
Product performance (ecommerce) — For campaigns linked to a Merchant Center feed, product-level performance data is available.
When to Use Performance Max
PMax works best when:
- You have a clear, measurable conversion goal (purchases, form fills, calls)
- You have conversion data in the account (50+ conversions per month ideally)
- You can provide high-quality creative assets including video
- You want to scale across multiple Google channels simultaneously
- You have an ecommerce store with a product feed
PMax is less suitable when:
- You're a new advertiser with little conversion history
- You need granular keyword-level control for brand safety
- You're advertising a product with restricted categories where automated placement is risky
- Your conversion volume is too low for the algorithm to learn efficiently
FAQs
Does Performance Max cannibalise other campaigns? It can. PMax typically takes priority over other campaign types for the same user in the same auction. Proper campaign segmentation and brand exclusions help minimise cannibalisation, but monitoring for overlap is important.
Can I add negative keywords to Performance Max? At a campaign level, only brand exclusions and account-level negative keyword lists apply directly to PMax. Campaign-level negative keywords are not available in the same way as standard campaigns, though Google has expanded this capability over time. Check current Google Ads documentation for the latest options.
How much budget does Performance Max need? PMax needs sufficient budget to exit the learning phase and gather conversion data quickly. As a general guideline, daily budget should be at least 10–20× your target cost-per-conversion. Under-funded campaigns extend the learning phase and underperform.
Is Performance Max worth using for B2B? For B2B with form-fill conversions and sufficient conversion volume, yes — but requires careful audience signal configuration and close monitoring of lead quality. The algorithm optimises for volume of conversion actions, not quality of leads.
For expert Google Ads management and PMax strategy, get in touch or start with a free audit.
