AI Max was sold as being a time saver for PPC managers. For some, that might be the case. Others claim it’s anything but.
Switch it on, let it do its thing and, according to Google, you could see a 14% uplift in conversions or conversion value at a similar CPA or ROAS. For campaigns still heavily reliant on exact and phrase match, that uplift reportedly climbs to 27%.
On the surface, it sounds like a win-win. More automation, less manual input and a helpful algorithm doing its thing in the background. But...come on, it was never going to be that simple, was it. The more time you spend with AI Max, the more you start to notice a different pattern emerging. It’s not necessarily doing less work - just not the bang-on-the-money work you need it to be doing.
AI Max in simple terms
AI Max works with your existing campaigns and enhances them with a bunch of automation features.
These include expanded search term matching, text customisation for ad copy and final URL expansion, where Google selects the most relevant landing page for each query (in a similar way to Dynamic Search Ads).
The idea is pretty simple. By using a wider range of signals, including intent, context and behavioural data, Google can match your ads to more searches than your keyword list alone would allow as well as embracing the new AI features within the SERPs. In practice, this can open the door to a level of unpredictability that many advertisers should be wary about.
The promise: more reach, better performance
To give Google its due, the benefits it highlights are not without merit.
AI Max can dig up new search queries that traditional keyword targeting might miss. It can adapt ad copy to better reflect user intent and, in some cases, direct traffic to more relevant pages on your site.
There’s also the potential for improved efficiency in the right circumstances. Accounts with strong conversion data and clear signals can give the system enough information to make reasonably good decisions, at least some of the time.
So yes, there are certainly scenarios where AI Max can be of significant value to PPC managers.
The problem is how much control you give up in the process.
Where things might start to go awry
Paid search has always been about striking the right balance. You define intent through keywords, curate messaging through ad copy and guide users through carefully selected landing pages. It’s not always perfect or easy, but it is deliberate.
AI Max relaxes that structure in three key areas, all at once.
1. Query matching becomes broader and less predictable
Google expands beyond your defined keywords using broad match and keywordless technology, relying on signals rather than strict targeting rules.
While that can surface useful queries, it also means your ads can appear for searches that only loosely relate to your offering. The system is designed to interpret intent, but intent can often be a grey area
2. Ad messaging is no longer entirely yours
With text customisation enabled, headlines and descriptions can be generated based on your site content, existing ads and keywords
Sometimes this results in relevant, well-structured ads. On other occasions, it produces messaging that feels off-brand or too generic. The challenge is that you’re no longer approving every variation before it goes live when it’s too late.
3. Landing pages are selected dynamically
Final URL expansion allows Google to choose which page a user lands on, depending on what it believes is most relevant to the query.
This can improve alignment in some cases. But if your campaigns are built around specific journeys or carefully optimised conversion paths, it can also disrupt the experience you intended to create.
Individually, these areas aren’t the end of the world. Together though, they introduce a level of uncertainty that impacts how campaigns need to be managed.
From proactive control to reactive fixes
Traditionally, PPC management is proactive. You define your targeting, refine your keywords and adjust your ads based on performance data. You hold the reins.
With AI Max, much of that upfront control becomes compromised. The system explores a wider range of queries and combinations, and you respond to what it finds. That response usually comes in the form of negative keywords. And this is where the extra legwork comes in.
The rise of the negative keyword workload
When AI Max pushes your ads towards less relevant territory, the only way to regain control is to actively exclude what you don’t want. At first, this might feel manageable. You review search terms, root out the iffy ones and add them to your negative list.
But as the system continues to explore new queries, more exclusions are needed. What might start off as the occasional bit of extra admin quickly snowballs into something bigger.
In one example we reviewed internally, a travel campaign that had previously been tightly controlled began appearing for searches including ‘cheap’, ‘disabled’ and ‘last minute’. All these terms were technically related to the broader category, but none matched the actual offering.
The end result was a lengthy list of exclusions and a daily laborious routine of monitoring and updating. It felt less like optimisation and more like whack-a-mole.
Why this approach is problematic
Negative keywords are inherently reactive. They address problems after they appear rather than preventing them in the first place. As campaigns expand and AI Max continues to test new queries, the volume of potential exclusions increases. Over time, this can lead to vast, complex negative keyword lists that are a major hassle to manage and maintain. The problem is further exacerbated by the fact that Google Ads accounts have caps on the number of negative keywords that can be applied, meaning even this reactive approach can eventually hit a ceiling.
More importantly, the time required to maintain these lists grows alongside them. Instead of reducing workload, AI Max can dump a whole lot more admin on your plate.
And that’s not all. There’s another challenge that is less obvious but equally important.
As AI Max expands matching, it becomes harder to clearly attribute performance. Independent analysis from Adalysis, reported by Search Engine Land, suggests that AI Max can blur match-type reporting and reassign impressions and clicks in ways that make it difficult to separate genuinely new traffic from existing demand.
This creates a visibility problem.
If you cannot clearly see which queries are driving results, it becomes harder to make confident decisions about where to invest budget. Optimisation slows down, and inefficiencies can persist for longer than they should.
It’s wise to be open-minded but cautious
AI Max isn’t necessarily a bad thing. It’s doing exactly what it was designed to do, which is to explore a broader range of opportunities and optimise towards performance goals. The challenge is that it operates without a full understanding of context or nuance.
PPC managers aren’t fearful of change. They’re sceptical because they’ve seen what happens when automation runs unchecked. Approach AI Max with an open mind, but also with caution. Dig into the data, track the trends and set boundaries before the system runs riot. Because at the end of the day, automation and expertise aren’t opposites. They’re partners - or at least, they should be.
