Google Ads AI Max Updates: What Advertisers Need to Know Before September 1

From 1 September, Google will begin automatically upgrading Search campaigns that still use campaign-level broad match or standalone automatically created assets (ACA) to AI Max for Search. The rollout happens gradually across the month, campaign by campaign, and once a campaign has been migrated there is no route back to the legacy setup.

If you run Search campaigns and haven’t actively looked at AI Max yet, this is the point to start. Not because AI Max is something every advertiser needs to switch on immediately, but because Google is about to make the decision for you on any campaign that still has these legacy settings turned on.

The timing happens to coincide with three genuinely useful developments Google has made to how AI Max can be tested and forecast. Multi-campaign experiments, the ability to keep brand and location controls in place during testing, and an expanded Performance Planner all give advertisers a more realistic way to evaluate AI Max before committing budget to it. Used properly, they’re the difference between adopting AI Max on your own terms and simply finding out what it does after the fact.

This article works through what AI Max actually is, why the September deadline matters, what’s changed in how you can test it, and a practical checklist for getting your account ready.

What Is AI Max for Google Ads?

AI Max for Search is Google’s expanded version of a standard Search campaign. Rather than relying solely on the keywords you’ve added, it uses AI to widen how your ads are matched to relevant search queries, and can generate additional headlines and descriptions based on your landing pages and existing assets.

In practical terms, AI Max builds on three things layered on top of a normal Search campaign:

• Search term matching, which lets Google match your ads to a broader range of queries than your keyword list alone would cover, in a similar spirit to broad match but with more automated interpretation of intent.
• Text customisation, where Google can generate or adjust ad copy using signals from your landing pages and creative assets.
• Final URL expansion, which allows Google to send traffic to a different page on your site than the one you specified, where it judges that page to be more relevant.

None of this is compulsory to switch on in full, and that’s an important point that tends to get lost. AI Max campaigns still have brand exclusions, URL exclusions and inclusions, and location controls available at the campaign and ad group level, so advertisers can decide how much latitude to give the system rather than accepting a single all-or-nothing setting.

For business owners, the simplest way to think about it: AI Max asks you to define the boundaries and the commercial goal, then gives Google more freedom to work within them. Whether that trade-off suits your account depends on your data, your margins and how tightly you need to control where your brand appears.

Why September 1 Matters for Google Ads Advertisers

Google has confirmed that, starting 1 September, Search campaigns using the campaign-level broad match setting or standalone automatically created assets will be automatically upgraded to AI Max. Google has said the migration will use equivalent AI Max settings intended to mirror the existing setup as closely as possible, and that existing brand inclusions and exclusions will carry over. Campaigns using campaign-level broad match will move across with search term matching switched on by default, without text customisation or final URL expansion enabled.

The only way to avoid the automatic upgrade on a given campaign is to turn off the affected legacy settings before the migration reaches that campaign. Google has also stopped advertisers creating new campaign-level broad match or legacy ACA setups altogether, which signals fairly clearly that AI Max is now the default direction for Search campaign management, not an optional add-on.

This creates a meaningful distinction. An advertiser who tests AI Max deliberately, understands how it behaves on their account, and decides to adopt it is in a completely different position to one whose campaigns are switched over automatically because nobody reviewed the account settings in time. The outcome for the campaign might end up looking identical either way. The difference is whether you made an informed decision or inherited one.

If you haven’t reviewed which of your Search campaigns are still running campaign-level broad match or ACA, that’s the first job before anything else in this article.

1. Multi-Campaign AI Max Testing Arrives in September

Until now, AI Max experiments have largely worked at the level of a single campaign, using Google’s one-click experiment structure to compare AI Max against a standard control. That’s useful, but it has an obvious limitation: a lot of commercial decisions aren’t really about one campaign in isolation, they’re about what happens when you change budgets or targets across several campaigns at once.

Google is rolling out the ability to run a single A/B test across multiple Search campaigns simultaneously. This means advertisers can test AI Max while adjusting budgets or ROI targets across a group of campaigns together, rather than having to piece together several separate single-campaign tests and guess at how they’d interact.

This matters most when you’re trying to understand:

• How AI Max performs across campaigns with different budget levels, not just your best-funded one
• Whether ROAS or CPA targets hold up consistently, or only in specific campaigns
• The account-level impact of a change, rather than an isolated result that might not represent the whole picture
• Whether scaling budget into AI Max campaigns produces proportional results or diminishing returns

An advertiser managing, say, six Search campaigns across different product lines or regions can get a genuinely more commercially meaningful read on AI Max this way than by testing it on a single campaign and extrapolating. That’s a real improvement, though it’s still a controlled test with its own limitations around sample size and seasonality, which we’ll come back to.

2. Brand and Location Controls Can Remain in Place During Testing

This is arguably the more practically important of the two testing changes, because it removes a genuine blocker for a specific group of advertisers. Previously, running an AI Max experiment could mean the test didn’t reflect the brand exclusions or location controls the business actually depends on, which made the results hard to trust, or in some cases commercially unacceptable to run at all.

Google’s AI Max experiments now support brand and location parameters directly. That means advertisers can test AI Max’s impact without having to strip out the exclusions and geographic restrictions that already matter to how they operate.

Why this matters varies by business type:

• Lead generation businesses often rely on brand exclusions to avoid paying for clicks from people searching for competitors, and location controls to stop budget being spent on enquiries from areas the business doesn’t service. Losing either during a test would make the results irrelevant to how the campaign would actually run.
• eCommerce brands may need to exclude marketplaces or resellers they don’t want to be associated with, or restrict targeting to countries they can actually fulfil orders in.
• Local businesses typically depend on tight location targeting as a core part of the campaign, not an optional extra, so a test that ignored it would tell them very little.
• B2B advertisers frequently exclude competitor brand terms and restrict campaigns to specific regions or account-based targeting lists, where any deviation changes who the campaign is actually reaching.

Being able to keep these guardrails in place during testing means the comparison you’re running is closer to a genuine like-for-like: AI Max versus your existing setup, under the same real-world constraints, rather than AI Max under looser conditions than you’d ever actually deploy.

3. Performance Planner Can Forecast the Impact Before Changes Go Live

The expanded Performance Planner lets advertisers model the potential effect of changes to budgets, bidding strategies or targets before making them live, and, where relevant, apply Google’s suggested changes directly from the forecast.

Used well, this gives you another input before committing spend, particularly useful when you’re deciding how much budget to move into an AI Max campaign or test, or what target to set it against.

It’s worth being clear about what this is and isn’t. A forecast is an estimate based on modelling, not a guarantee of what will actually happen once a campaign goes live. Auction dynamics, seasonality, competitor behaviour and changes elsewhere in the account can all move actual results away from a forecast. Performance Planner is best treated as one data point that informs a decision, sat alongside your own historical performance data and commercial judgement, rather than something to follow uncritically.

What Should Advertisers Do Before September 1?

This is the part that matters most in practice. A simple four-step framework covers the essentials.

1. Audit campaigns still using standard broad match at campaign level. Go through your account and identify exactly which Search campaigns are running campaign-level broad match or standalone ACA. For each one, note current performance so you have a baseline to compare against, whatever happens next.

2. Run controlled multi-campaign AI Max experiments. Rather than committing a large share of the account or budget straight away, use the new multi-campaign testing capability to see how AI Max performs across a representative group of campaigns first.

3. Keep brand and location guardrails in place. Set up experiments using the same brand exclusions and location controls you’d actually want live, so the results reflect the conditions the business genuinely operates under, not a loosened version of them.

4. Use Performance Planner before changing budgets. Before moving spend into AI Max campaigns or adjusting targets, forecast the likely impact first, and weigh that against your own account history.

The point of all four steps together is to reach 1 September having made a deliberate decision about your AI Max strategy, rather than reacting once campaign behaviour has already changed on its own.

What Should You Measure During an AI Max Test?

Clicks and impressions won’t tell you much on their own. The metrics that actually indicate whether AI Max is working for your account go further:

Conversion volume, cost per acquisition, conversion value and ROAS remain the core commercial measures, but they need to be read alongside lead quality, search term relevance, incremental conversions, geographic performance, brand versus non-brand performance, and budget utilisation.

For lead generation businesses in particular, it’s worth stating plainly: more conversions is not automatically better performance. If AI Max’s broader matching brings in a higher volume of leads but a larger share of them are poor fits, unqualified, or simply don’t convert to sales further down the funnel, the campaign may look better in Google Ads while actually performing worse for the business. Sales team feedback and CRM data on lead quality matter as much as the numbers inside the ad platform.

AI Max Doesn’t Mean Giving Google Unlimited Control

AI Max sits within a broader shift in how Search advertising works, where more of the day-to-day optimisation is handled by automated systems rather than manual keyword and bid management. That shift isn’t inherently good or bad. It depends heavily on what’s feeding it.

Automation performs well when it’s built on accurate conversion tracking, clean and sufficient data, sensible exclusions, clearly defined commercial targets, and ongoing human oversight. It performs poorly when any of those are missing, because the system ends up optimising towards the wrong signal or without the guardrails a business actually needs.

Practically, this changes what a PPC specialist spends their time doing. Less time goes into manually adjusting individual keyword bids, and more goes into designing the framework the automation operates within: what data it’s fed, what exclusions and controls are set, what the targets actually are, and how results are checked against business outcomes rather than platform metrics alone. The skill hasn’t disappeared. It’s moved up a level.

Should You Switch to Google Ads AI Max?

There’s no single right answer here, and anyone telling you there is hasn’t looked closely enough at your account.

The right decision depends on a combination of factors: how your existing campaigns are performing, the quality and completeness of your conversion tracking, how mature the account is, the search volume available in your market, your budget, your wider business objectives, the quality of the leads or customers you’re generating, and your ability to measure revenue or other downstream outcomes rather than just platform-reported conversions.

An account with strong conversion tracking, healthy search volume and clear commercial targets is in a good position to test AI Max properly and learn something useful from it quickly. An account with patchy tracking or very low volume may struggle to get a reliable read either way, which is itself useful information.

Controlled experimentation, using the tools covered above, is the sensible route in almost every case. Blindly enabling AI Max because it’s the new default, and blindly avoiding it because it’s unfamiliar, both skip the step that actually matters: finding out how it performs on your account, with your data, against your targets.

Google Ads AI Max Checklist Before September 1

• Identify affected Search campaigns
• Review conversion tracking
• Review brand exclusions
• Review geographic targeting
• Establish CPA/ROAS guardrails
• Launch controlled experiments
• Compare lead/customer quality
• Review Performance Planner forecasts
• Document baseline performance
• Decide your AI Max strategy before September 1

Final Thoughts

The additional AI functionality isn’t really the headline here. What matters more is that Google is giving advertisers better tools to test, forecast and maintain control at exactly the point Search advertising is becoming more automated by default.

Businesses that use the weeks before 1 September to test deliberately, on their own terms, will be in a stronger position than those that wait to see what happens once the migration reaches their account. The campaigns are moving to AI Max either way. The only real choice is whether that happens with your input or without it.

Need Help Preparing Your Google Ads Account for AI Max?

Mass Reach is a Google Partner performance marketing agency helping UK B2B and eCommerce businesses manage, test and optimise Google Ads around measurable commercial growth.

If you’re unsure how AI Max could affect your campaigns, we can audit your existing Google Ads account, identify which campaigns may be affected, and build a controlled testing strategy before you commit any further budget. This sits alongside our wider Performance Max management, Google Shopping Ads and conversion rate optimisation work, and you can see examples of results we’ve delivered for other clients in our case studies.

Book a Free Google Ads Strategy Call