One of the hardest questions in Google Ads is also one of the simplest: “What happens if we spend more?” On 20 August 2026, Google announced new testing and planning tools for Search and AI Max campaigns that are designed to answer exactly that. Multi-campaign A/B tests began rolling out in September, alongside changes to AI Max experiments and Performance Planner.
With Google moving so much of Search onto AI Max this year, better testing tools are welcome. Here’s what’s been announced and how a small or mid-sized business can use them without needing a data science team.
1. Multi-campaign A/B tests
Until now, most Google Ads experiments worked one campaign at a time. That’s fine for testing an ad or a bid strategy, but it doesn’t reflect how businesses actually grow. When you scale, you usually change budgets or targets across several campaigns at once.
Google’s new multi-campaign experiments, rolling out from September 2026, let you test different budgets and ROI targets across multiple Search campaigns in a single A/B test. As Search Engine Journal put it, advertisers can test a broader strategy while keeping a control group for comparison.
For example, a business might test whether lifting budgets and relaxing target ROAS across all its non-brand campaigns brings in enough extra sales to justify the spend, rather than guessing from before-and-after numbers that are muddied by seasonality.
2. AI Max experiments that keep your guardrails on
Previously, if you wanted to test AI Max, brand and location controls had to be removed for the experiment. That meant the test didn’t reflect how the campaign would really run, which made the results hard to trust.
Now, AI Max experiments support brand and location controls. In Google’s words, advertisers can test the impact of AI Max without compromising those guardrails. That’s particularly relevant for:
- local service businesses that only work in certain areas
- resellers and franchisees with rules about which brand terms they can bid on
- businesses that want to keep competitor brand searches out of their campaigns
3. Performance Planner forecasts you can apply in one click
Performance Planner is Google’s forecasting tool. It now projects how changes to bidding or budget targets may affect the performance of existing campaigns, and lets you apply Google’s suggested changes directly with one click. According to Search Engine Journal, those changes appear in Bulk Actions and can be undone if needed.
A word of caution: a forecast is a model, not a promise. Performance Planner is useful for spotting where there may be room to grow, but we’d always pair a big change with a proper experiment rather than applying it across the account on day one.
Why this matters right now
Throughout September, Google has been automatically upgrading Search campaigns that use automatically created assets or campaign-level broad match to AI Max. Many advertisers now have AI Max running whether they chose it or not. These testing tools give you a fair way to check whether it’s working for your business, and whether scaling up is worth it.
It also matters because the end of the year is when many Australian businesses are tempted to push budgets harder. Running a proper test in the next few weeks gives you evidence to back that decision, rather than simply turning everything up and hoping the extra spend pays off. And if the test shows that extra budget mostly buys more expensive clicks, you’ve saved money you can put somewhere more productive.
What to do now
- Write down the question first. “Will a 30% higher budget across non-brand Search deliver leads at an acceptable cost?” is testable. “Is Google working?” isn’t.
- Check your conversion tracking before testing. An experiment is only as good as the data it measures. Make sure you’re counting real leads or sales, ideally with values attached.
- Use AI Max experiments with your normal controls switched on. Keep brand and location settings as they would be in real life, so the result reflects reality.
- Give tests enough time and budget. Low-volume accounts may need several weeks to reach a meaningful result. Avoid calling a winner after a few days.
- Treat Performance Planner as a starting point. Use it to find opportunities, then validate the biggest ones with an experiment.
- Keep a change log. Note when tests start and finish so you can explain changes in results later.
Good testing is what separates confident growth from expensive guesswork. If you’d like help designing a Google Ads experiment that answers the questions that matter to your business, the Big Digital team is happy to help.
Photo: Luke Chesser on Unsplash