Google Ads AI campaigns: Performance Max and AI Max explained

How Google's AI-driven campaign types work, how ads now appear in AI Overviews and AI Mode, and how to feed the algorithms the right signals to protect your budget.

By the UK Technology Hub teamPublished 4 minute read

The short answer

Google Ads increasingly relies on AI to decide where, when and to whom ads appear. Performance Max runs across Google's channels from one campaign, and AI Max for Search expands search campaigns beyond exact keywords, including into AI Overviews and AI Mode. To succeed, feed the system accurate conversion data and values, strong creative and product data, sensible guardrails and clear goals, then judge results on profit.

Paid search used to be about choosing keywords and setting bids. Google's AI now does much of that work. Advertisers who understand how to guide the algorithms get strong results; those who hand over control without good inputs can waste budget quickly.

Features and names change frequently, so check Google's current documentation for details.

The main AI-driven campaign types

Performance Max (PMax). One campaign that runs across Search, Shopping, YouTube, Display, Discover, Gmail and Maps. You provide goals, budgets, creative assets, audience signals and product feeds; Google's AI decides placements and bids.

AI Max for Search. An option for Search campaigns that uses AI to match ads to relevant searches beyond your keywords, tailor ad text and choose landing pages, while keeping more transparency and control than PMax.

Smart bidding. Automated bidding strategies, such as target cost per acquisition or target return on ad spend, used across campaign types.

Ads in AI Overviews and AI Mode

Google has expanded ads into AI Overviews and AI Mode, where they appear labelled as sponsored. Eligibility generally comes through campaign types such as Performance Max, Shopping, broad match Search and AI Max for Search. As AI answers occupy more of the results page, these placements may become an important part of search visibility.

What the algorithms need from you

Accurate conversion tracking

AI optimises for whatever you tell it is a conversion. Track meaningful actions, such as purchases, qualified leads and booked calls, not just page views or button clicks. Use enhanced conversions and offline conversion imports where possible, so the system learns which leads become customers.

Conversion values

If some conversions are worth more than others, pass values so the system optimises for revenue or profit, not volume.

Quality creative and assets

Provide strong headlines, descriptions, images and videos. Poor assets limit performance across every placement.

Product data

For e-commerce, an accurate, detailed product feed with good titles, images and prices is essential. Our e-commerce team helps keep feeds clean.

Audience signals

Customer lists and interest signals help the AI find similar people faster.

Guardrails to set

  • Brand exclusions and negative keywords where available.
  • Budget limits while testing.
  • Location and language settings.
  • Landing page controls so ads don't send people to irrelevant pages.
  • Brand safety settings for video and display placements.

Measuring what matters

Platform-reported conversions can include modelled data and overlap with other channels. Compare against your own sales data, run incrementality tests where possible and look at profit after ad costs. Our marketing attribution guide explains approaches.

Testing changes properly

Test one significant change at a time, such as enabling AI Max or adding new assets, and compare results over a meaningful period. Google offers experiments for some campaign types. Our A/B testing guide covers testing principles.

Landing pages still matter

AI can send more traffic, but your website has to convert it. Fast, clear landing pages with relevant offers improve both conversion rates and ad performance. Our web development team builds landing pages designed for paid traffic.

Common mistakes

  • Optimising for low-quality conversions such as page views.
  • Launching PMax without exclusions, so it bids heavily on branded searches you'd win anyway.
  • Weak or few creative assets.
  • Judging success by platform metrics alone.
  • Making frequent changes before campaigns have time to learn.

A worked example

A home improvement company switched to an AI-driven campaign and saw leads rise, but many were outside its service area or asked for jobs it didn't do. The fix wasn't abandoning automation. It tightened location settings, added negative keywords, imported offline conversions that showed which leads became paid jobs and gave the system better images and descriptions. Within weeks, lead volume settled slightly lower, but far more became customers.

Keeping humans in charge

Review search terms, placements and asset reports at least weekly. Automation optimises for the signals you give it, so poor signals produce poor results quickly.

How we help

Our AI solutions team connects CRM and sales data to advertising platforms for better optimisation, our web team builds high-converting landing pages and our e-commerce team keeps product feeds accurate.

Frequently asked questions

Is Performance Max good for small businesses?

It can be, with accurate conversion tracking, good assets and sensible budgets. Without these, results can be poor.

Can I advertise in Google's AI Overviews?

Ads can appear in AI Overviews and AI Mode through eligible campaign types. Check Google's current eligibility rules.

Do I still need keywords in Google Ads?

Keywords still guide Search campaigns, but AI features increasingly match ads to related searches beyond them.

This guide is general information, not legal, tax or financial advice, and details such as prices, features and rules can change. Product and company names mentioned are trademarks of their respective owners and are used only to identify those products; UK Technology Hub is not affiliated with or endorsed by them.

Related guides

Marketing attribution in a privacy-first world

How to understand which marketing actually drives results when tracking is limited, combining attribution models, experiments and marketing mix modelling.

All digital marketing guides