.png)
Let’s be honest: Change is the only constant in advertising. It keeps it exciting, but it also keeps you on your toes.
Here’s why: Big ad tech platforms like Google are steadily taking control out of advertisers’ hands, replacing years of manual decision-making with automation and AI.
So pay attention and know what you’re agreeing to when receiving “free” upgrades and not-so-subtle pushes from reps. Because, at the same time, regulators are raising questions about whether these platforms can be trusted to grade their own homework.
The shift from advertiser-stewarded to platform-stewarded buying has only accelerated. In September, Google began automatically moving Search campaigns that already used campaign-level broad match or auto-generated assets into AI Max. Google is also retiring standalone Display campaigns and folding them into its Demand Gen product.
And, beginning October 12th, Google has been automatically adding sales and promotions it finds on advertisers’ websites to eligible ads unless they opt out, a risky default heading into peak season. Microsoft has since followed suit: advertisers can no longer set a maximum cost per click on new campaigns that use automated bid strategies like Target CPA and Target ROAS. Existing campaigns keep the cap, but once you remove it, there’s no turning back.
This even quieter change is even more controversial. Instead of maintaining performance when campaigns outperform their targets, Google now steers results closer to the targets advertisers set. So, if your campaign has a $10 CPA target but has consistently been delivering conversions at $5, Google will likely optimize for the $10 target rather than maintain the lower cost. One analysis found that after the change, median cost per click in budget-limited Target ROAS campaigns rose 15.8%.
Why is this big drama? Your target is no longer treated as a ceiling. It’s becoming the number Google aims to hit, even when your campaign could perform better.
To be fair, Google has handed some control back. In mid-October, it launched channel prioritization settings that let advertisers rank and cap budget by channel, and it is testing a report that shows which headlines and landing pages AI Max serves for specific searches. Both are steps in the right direction, but there’s still a catch: Google is the one reporting on how well Google is performing.
On August 31, the FTC and attorneys general from 22 states sued Amazon. They allege that for more than seven years, Amazon inflated what advertisers, including more than 500,000 small and medium-sized businesses, paid for Sponsored Products, Sponsored Brands and Sponsored Display ads, partly through what the complaint describes as an “invented auction participant.” The FTC estimates the hidden surcharges cost advertisers more than $20 billion. Amazon denies the allegations, saying average cost per click stayed flat from 2019 to 2024 while the sales those clicks generated rose – we’ll let the judges decide.
Regardless of the verdict, the case represents a universal problem: Every major platform runs an auction you can't see inside, then grades its own results. As platforms automate more and disclose less, that self-graded scorecard becomes the only view many teams have of their own spend.
The answer isn’t to reject automation. In many cases, it outperforms manual management. The key is to be deliberate about where and how you use it, while putting the right guardrails in place to protect your brand and ensure accountability.
To do this, every brand needs to have a handle on its data and a POV on how best to move forward.
Audit platform targets against your own company’s economics. Compare how targets for KPIs like CPA and ROAS across platforms hold up against your current customer value.
Reconcile platform numbers with in-house data. Compare platform-reported conversions and revenue against internal financial systems to understand the gap and track if it grows.
Fund incrementality testing to continue to test and learn what ad spend is actually driving, not just touching. Google’s Conversion Lift testing is now self-serve for Search and Performance Max in qualifying accounts, which lowers the barrier to getting started.
Decide platform by platform what makes sense for your business. What buys has automation helped dramatically? Which platforms can you not verify results for? Make sure the direction you move forward with is intentional and data-driven.
Stay ahead of platform updates so you can maintain control over campaign performance. Every business has critical sales periods when results matter most. Be proactive about monitoring changes and making adjustments ahead of these key windows to minimize disruption and protect performance.
TLDR
Automation is not the enemy, and platforms aren’t the villains – but the AI landscape is forcing us to adapt how we think and act as marketers. The teams that thrive are the ones that are critical of platform outcomes and can compare them to internal data to prove advertising value, rather than relying on platforms to toot their own dashboards.
.png)
