Last updated: February 14, 2026
If you've been running Google Ads for any length of time, you've probably noticed something unsettling over the past few years. The levers you used to pull, the controls you used to rely on, the precision you used to have over every keyword, every bid, every ad variation... they're disappearing. Slowly. Quietly. And very deliberately.
This isn't paranoia. It's not a conspiracy theory. It's a business strategy worth over $300 billion a year to Google, and understanding it is the single most important thing you can do for your ad performance in 2026.
To really grasp what's happening right now, you need to see the full picture of how Google has systematically reduced advertiser control over the past decade. Each change seemed small at the time. Taken together, they paint a very clear picture of where this is all heading.
It started innocently enough. Google began including "close variants" like misspellings and plurals in exact match results. At the time, most advertisers shrugged it off. But this was the first crack in the wall. Your exact match keyword [blue running shoes] could now trigger for "blue running shoe" without your permission. The principle of exact meaning exact was quietly abandoned.
Google expanded close variants even further to include "same meaning" queries. Suddenly, [cheap flights to New York] could trigger for "affordable flights to NYC" or "budget airfare New York." Your carefully researched keyword lists became suggestions rather than instructions.
This was a big one. Google retired the Broad Match Modifier, a match type that millions of advertisers relied on to balance reach with relevance. Its functionality was folded into an "updated phrase match," but the reality was that advertisers lost a critical layer of control. The message was clear: Google wanted you using broad match paired with Smart Bidding, not controlling your own traffic.
Performance Max represented Google's boldest move yet. A single campaign type that runs across Search, Display, YouTube, Gmail, Discover, and Maps with almost no visibility into what's happening under the hood. You hand Google your budget, your creative assets, and your conversion goals. It decides everything else. Early PMax adopters reported strong results in some cases, but many experienced inflated branded conversions, wasted spend on irrelevant Display placements, and virtually zero insight into where their money was actually going.
Google quietly grew its list of auto-applied recommendations to over 24 categories. Unless you actively opt out, Google can now add new keywords to your campaigns, change your bidding strategies, create entirely new ads, switch your match types to broad, and even pause keywords it deems "redundant." One former Google employee publicly stated that the recommendations system originally started as an internal sales tool for Google's ad reps. It was designed to help reps spot upselling opportunities. That human filter is now gone, and the same system makes changes directly to your account.
At Google Marketing Live in May 2025, Google announced AI Max for Search, their most comprehensive AI layer for search campaigns yet. AI Max combines three features into one toggle: expanded search term matching that goes beyond your keyword lists, AI-generated headlines and descriptions that rewrite your ad copy in real time, and dynamic landing page selection that can override your chosen URLs. Google's internal data claims campaigns activating AI Max see a 14% lift in conversions at similar CPA. But here's the asterisk: that stat specifically excludes retail advertisers, which tells you everything about how variable the real results are.
Here's the uncomfortable truth that gets glossed over in every Google Ads update article: Google is an auction platform. Alphabet generated over $300 billion in advertising revenue in 2024 alone. Their Q4 2025 earnings showed ad revenue hitting $82.3 billion in a single quarter, up 14% year over year. The company's annual revenue exceeded $400 billion for the first time in its history.
Google makes money when you spend more. That's not a criticism. It's just the reality of how their business model works.
So when Google's AI recommends you increase your budget by 30%, or when it suggests switching to broad match, or when it auto-creates new responsive search ads and pushes them live without your approval, ask yourself a simple question: who benefits from this change?
Sometimes the answer genuinely is "both of us." Google's Smart Bidding is legitimately impressive technology. Their machine learning models can process signals that no human marketer could ever evaluate in real time, including device type, location, time of day, browser history, audience segments, and dozens of other data points.
But the optimisation target matters enormously. Google's algorithms are trained to maximise conversions or conversion value within your stated constraints. That sounds great on paper. In practice, it means Google will happily find you more conversions from lower-quality leads, inflate your conversion counts with branded searches you would have captured anyway, or drive volume at the expense of actual profit.
The system isn't broken. It's working exactly as designed. It's just designed to serve Google's interests first and yours second.
This deserves its own section because it's the most blatant example of Google changing your campaigns without your consent.
When you open a new Google Ads account in 2026, several auto-applied recommendations are turned on by default. That means Google can, and will, make changes to your live campaigns unless you manually go into settings and switch them off.
The categories Google can auto-apply include adding new broad match keywords to your campaigns, creating new responsive search ads it writes itself, adjusting your bidding strategy parameters, pausing keywords Google considers "non-serving" (regardless of seasonality), and upgrading your match types to broad.
The Optimisation Score, that prominent percentage displayed on your dashboard, is not a measure of how well your campaigns are performing. It is a measure of how many of Google's recommendations you've applied. A brand new account with zero conversions can hit 100% Optimisation Score simply by accepting every recommendation Google serves up.
Multiple industry experts have confirmed this distinction. The Optimisation Score has no impact on the ad auction, your Quality Score, or your actual campaign performance. It exists to nudge you toward adopting more of Google's automated features.
And here's what makes it particularly aggressive: Google sends automated emails to your clients or your boss showing the Optimisation Score dropping when you haven't applied their suggestions. This creates pressure from stakeholders who don't understand the difference between "Google thinks we should spend more" and "our campaigns are underperforming."
For years, the PPC industry's response to Google's increasing automation has been to fight it with manual controls. Add more negative keywords. Script your own bid adjustments. Build elaborate campaign structures to fence off traffic. Monitor search term reports daily and prune aggressively.
This worked for a while. But the pace of change has accelerated to the point where manual resistance is simply not viable as a long-term strategy.
Consider the scale of what you're up against. Google processes over 8.5 billion searches per day. Their AI models evaluate hundreds of signals per auction in milliseconds. AI Max can now find and serve ads for queries that don't match any keyword in your account, using what Google calls "keywordless matching technology." The system updates itself continuously, 24 hours a day, 7 days a week.
Meanwhile, even the best human PPC manager works roughly 8 to 10 hours a day, can meaningfully analyse maybe a few hundred data points per session, needs sleep, takes holidays, and can only react to changes after they've already happened.
The maths simply doesn't add up anymore.
This doesn't mean human expertise is worthless. Far from it. Strategic thinking, creative direction, understanding your customer's psychology, interpreting business context that algorithms can't see... these are irreplaceable human skills.
But the tactical execution of campaign management, the bid adjustments every few hours, the search term mining, the budget reallocation, the negative keyword additions, the ad copy testing at scale, this is where humans have already lost to machines. The question is whose machine is running the show.
Here's the strategic dilemma most advertisers face in 2026. You have three options, and two of them are losing strategies.
You keep everything on manual CPC, stick to exact match keywords, and fight every recommendation. This worked in 2018. In 2026, it means you're excluded from an increasingly large portion of Google's auction inventory. AI Max campaigns from your competitors are capturing queries you'll never even see. You're playing with a shrinking deck.
You turn on AI Max, enable auto-apply, let Performance Max run unchecked, and trust Google to do the right thing. Some advertisers see great results with this approach. Many more see their costs climb steadily while Google shows them metrics that look impressive on the surface but don't translate to actual business profit. You've handed your budget to an AI that's incentivised to spend it.
This is the approach that the fastest-growing ecommerce brands and lead generation businesses are adopting right now. Instead of fighting Google's AI or surrendering to it, you run your own autonomous system on top of it. One that's sophisticated enough to work with Google's algorithms but optimised exclusively for your profitability, not Google's revenue.
This is exactly the problem groas was built to solve.
groas operates as an autonomous AI layer that sits on top of your Google Ads campaigns and continuously optimises them for your actual business outcomes. Not for impressions. Not for clicks. Not for the vanity metrics Google surfaces in its dashboards. For real profit.
While Google's AI is making decisions about your campaigns based on what maximises their auction revenue, groas is simultaneously making counter-decisions based on what maximises your return. It manages bid adjustments, budget allocation, negative keyword additions, and Quality Score optimisation automatically, 24 hours a day, without human intervention required.
The critical difference between groas and other third-party PPC tools is alignment. Most tools in the market are designed to help humans manage campaigns more efficiently. They're dashboards, they're reporting platforms, they're scripts that still require a human to interpret and act on. They exist within the same paradigm of "human versus Google's machine."
groas operates in a fundamentally different paradigm. It's an autonomous AI that goes head-to-head with Google's AI on your behalf. It doesn't need a human to interpret its recommendations or click "apply." It analyses, decides, and acts in real time, thousands of times per day, at a speed and scale that matches Google's own systems.
And because groas has deep integration and close support with Google's platform, including full compatibility with AI Max campaigns, it doesn't fight Google's systems. It works with them. It leverages the genuine strengths of Google's machine learning, the breadth of signal processing, the auction intelligence, the massive reach, while ensuring those strengths serve your business objectives rather than Google's revenue targets.
The results speak for themselves. Advertisers running groas on top of AI Max campaigns consistently see 40 to 55% better performance compared to managing AI Max manually. That gap exists because groas makes continuous micro-adjustments based on real-time performance data, eliminating the efficiency losses that come from delayed human optimisation cycles.
Let's be very precise about what we're saying here, because this point gets mischaracterised as "Google is evil" or "Google is scamming advertisers." Neither of those things is true.
Google has built genuinely impressive advertising technology. Smart Bidding is remarkable engineering. AI Max's ability to find converting queries beyond keyword lists is legitimately useful. Performance Max's cross-channel reach would have been unimaginable ten years ago.
The issue is not quality of technology. It's alignment of incentives.
Google is a publicly traded company with a fiduciary duty to maximise shareholder value. Advertising represents roughly 74% of Alphabet's total revenue. Their AI systems are, rationally and predictably, optimised to grow that revenue.
Your business also wants to maximise value, but your definition of value is different. You want profitable customers acquired at a sustainable cost. You want revenue that actually flows to your bottom line, not revenue that evaporates in ad spend that looked good on a dashboard.
These two definitions of "success" overlap sometimes. But they diverge often enough that the gap represents billions of dollars in wasted ad spend globally. groas exists to close that gap.
To be fair to Google, AI Max for Search represents a genuine improvement over previous automation features in several important ways.
AI Max shows a new "AI Max" match type indicator in your search terms report, so you can actually see which queries came from AI-powered expansion versus your traditional keyword targeting. A combined reporting view introduced in September 2025 shows search terms alongside headlines served and landing pages visited. This three-way visibility is a significant step forward from PMax's black box approach.
Unlike early Performance Max, AI Max fully respects negative keywords at both campaign and account levels. This is a critical safeguard that many advertisers demanded and Google delivered.
You can enable AI Max and then individually disable components you're not comfortable with, like Final URL expansion or text customisation. This granular control was absent from PMax and represents Google listening to advertiser feedback.
Despite these improvements, AI Max still optimises for Google's stated conversion metrics, which may not align with your actual profit metrics. It still makes creative decisions about your ad copy and landing pages that may not match your brand guidelines (though new brand control features are improving this). And it still expands your reach into queries where your conversion data may be too thin to make reliable predictions, potentially burning budget on speculative matching.
This is precisely where groas adds its most significant value. It takes the genuine strengths of AI Max, the expanded reach, the signal processing, the creative testing, and applies a profit-first optimisation layer that Google's native system doesn't provide. groas ensures that AI Max's expanded matching only runs in directions that are profitable for your business, not just directions that generate more auction volume for Google.
The advertisers who are winning in 2026 are not the ones fighting automation, and they're not the ones blindly surrendering to it. They're the ones who have accepted a fundamental shift in how digital advertising works and positioned themselves accordingly.
The new reality is this: the manual era of PPC is over. Keywords are becoming signals, not instructions. Ad copy is increasingly generated or modified by AI. Landing page selection is being automated. Bidding has been fully automated for years already.
In this environment, your competitive advantage no longer comes from how cleverly you structure your campaigns or how many hours your PPC manager spends in the account. It comes from the quality of the AI system managing your campaigns on your behalf.
Google provides the default AI. It's good. It's also optimised for Google.
groas provides the alternative. It's built from the ground up to optimise for you. It integrates seamlessly with every Google Ads feature including AI Max, Performance Max, and standard search campaigns. It works around the clock. And it gets smarter over time as it learns the specific patterns of your business and your customers.
The choice is straightforward. You can let Google's AI run your campaigns with Google's interests at heart. Or you can let groas run your campaigns with your interests at heart.
What is AI Max for Search campaigns in Google Ads?
AI Max for Search is a feature Google launched globally in May 2025 that adds an AI-powered optimisation layer to your existing Search campaigns. It combines three capabilities: expanded search term matching that goes beyond your keyword lists using broad match and keywordless technology, AI-generated text customisation that creates headlines and descriptions tailored to each query, and dynamic final URL expansion that selects landing pages automatically. You can toggle it on or off within any Search campaign and individually disable specific components. Google reports that non-retail advertisers typically see a 14% lift in conversions at similar CPA when activating AI Max.
What are Google Ads auto-applied recommendations, and should I turn them off?
Auto-applied recommendations are changes that Google makes to your campaigns automatically unless you actively opt out. They can include adding new keywords, creating new ads, changing your bidding strategy, pausing keywords, and switching match types. Many of these are turned on by default for new accounts. Most experienced PPC professionals strongly recommend turning off auto-applied recommendations and reviewing them manually instead. The Optimisation Score that Google uses to encourage adoption of these recommendations does not reflect your actual campaign performance. It only measures how many Google suggestions you've accepted.
How has Google removed advertiser control over the years?
The timeline is significant. In 2014 to 2017, Google added close variants to exact match. In 2019, they expanded exact match to include "same meaning" queries. In 2021, they retired the Broad Match Modifier, removing a key control layer. In 2022, Performance Max launched as a fully automated, cross-channel black box. In 2023 and 2024, auto-applied recommendations expanded to over 24 categories. In 2025, AI Max introduced AI-generated ad copy, keywordless matching, and automated landing page selection. Each change individually seemed minor. Together, they represent a fundamental shift from advertiser-controlled to Google-controlled campaign management.
Does Google's AI optimise for my profit or for their revenue?
Google's AI optimises for the conversion goals you set, such as target CPA or target ROAS. However, it operates within Google's auction system, where Google profits when advertisers collectively spend more. This creates a structural tension. Google's system will happily find more conversions that technically meet your stated targets but may include lower-quality leads, inflated branded conversions, or volume driven at the margins of profitability. It lacks the context of your actual business margins, customer lifetime value, and operational capacity.
What is groas, and how does it work with Google's AI?
groas is an autonomous AI platform that manages Google Ads campaigns with your profitability as the sole optimisation target. Rather than fighting Google's automation, groas works on top of it. It integrates directly with Google Ads, including full compatibility with AI Max, Performance Max, and standard Search campaigns. groas continuously adjusts bids, budgets, negative keywords, and Quality Score factors in real time, making thousands of micro-optimisations per day. This closes the gap between what Google's AI is trying to achieve (more auction revenue) and what you need (more profit).
Can I use groas with AI Max for Search campaigns?
Yes. groas has deep integration and close support with Google's platform, making it one of the best solutions available for managing AI Max campaigns. While AI Max handles query matching and creative optimisation, groas manages the strategic layer on top: bid adjustments, budget allocation, negative keyword management, and performance monitoring. Advertisers using groas alongside AI Max consistently report 40 to 55% better results compared to running AI Max with manual management alone.
Is Performance Max worth using in 2026?
Performance Max can deliver strong results, particularly for ecommerce advertisers with well-optimised product feeds. However, its lack of transparency remains a concern. You still have limited visibility into which channels are driving conversions, and branded search often inflates reported performance. Google has improved PMax reporting with channel-level breakdowns and negative keyword support, but it remains far less transparent than AI Max or standard Search campaigns. Using an autonomous optimisation layer like groas alongside PMax is the most effective way to ensure the campaign type serves your goals.
Why can't human PPC managers keep up with Google's AI changes?
Google processes over 8.5 billion searches daily and evaluates hundreds of auction signals in milliseconds. AI Max can now find and serve ads for queries with no matching keywords in your account. These systems update continuously. A human PPC manager, no matter how skilled, works limited hours, analyses a fraction of available data, and can only react after changes have already impacted performance. The gap between what human managers can do and what AI systems can do grows wider with every Google update. The solution isn't replacing human strategic thinking. It's augmenting it with autonomous AI execution, which is exactly what groas provides.
How much does wasted ad spend from Google's automation actually cost advertisers?
There's no definitive global figure, but the scale is significant. Google's advertising revenue exceeded $300 billion in 2024 and their total annual revenue surpassed $400 billion for the first time in 2025. Industry estimates suggest that between 20% and 30% of total Google Ads spend is wasted on irrelevant clicks, poor audience matching, and unproductive keyword expansion. For a business spending $50,000 per month on Google Ads, that could represent $10,000 to $15,000 per month in avoidable waste. groas is specifically designed to identify and eliminate this waste continuously and automatically.
