Key Takeaways
- AI-native ad tools built by Google, Meta, and Amazon are designed to keep your budget inside their ecosystems.
- Brands that hand full control to automated campaign tools often lose visibility into where their money actually goes and why.
- A smart media mix strategy means entering them on your terms, with your data intact.
- First-party data is your most valuable negotiating asset with Big Tech platforms, and most CPG brands aren’t using it aggressively enough.
- The efficiency gains from AI-powered buying are real, but they accrue disproportionately to the platforms, not the advertisers, unless you structure campaigns deliberately.
- Brands that diversify across emerging channels now will have more leverage when Big Tech eventually raises the cost of that AI-driven efficiency.
—
AI ad spend walled gardens are the current default. The moment a brand turns on Google’s Performance Max, Meta’s Advantage+ Shopping, or Amazon’s automated campaign tools, they are handing the keys to algorithms optimized for platform revenue, not brand growth. That distinction matters, and most brands aren’t thinking clearly about it.
Why AI-Powered Ad Tools Favor the Platform, Not the Advertiser
Google, Meta, and Amazon are building AI tools that are genuinely impressive at generating short-term conversion signals. But every one of those tools has one primary objective: maximize inventory fill and platform revenue. Your ROAS goal is secondary.
When you opt into full automation on any of these platforms, the system decides where your ad appears, who sees it, what creative runs, and what counts as a conversion. You get a dashboard, not transparency.
We have written about this dynamic specifically with Amazon in our breakdown of Amazon’s ad spend transparency problem. The issue is what you give up when you let it run unchecked.
The CPG and DTC Brand Caught in the Middle
For CPG and DTC brands, this concentration risk is acute. You are operating on thin margins, fighting for shelf space and search placement simultaneously, and trying to build brand equity while hitting short-term sales targets.
AI-powered tools can help with that last part. They are genuinely good at finding in-market signals and converting them. The problem is they do it by bidding more aggressively within the platform’s own ecosystem, which raises CPMs for everyone and pushes weaker-capitalized brands out of the auction.
Bigger budgets win the AI game on Big Tech platforms. That isn’t a level playing field for a growing organic snack brand or a DTC personal care company.
What “Efficiency” Actually Costs You
Here is the math brands often miss. An AI-optimized campaign on Performance Max might deliver a better in-platform ROAS than a manually structured campaign. But if you can’t see which placements drove that performance, you can’t replicate it, negotiate against it, or apply those learnings anywhere else.
We see this play out with clients who come to us after running fully automated campaigns for 12 to 18 months. The numbers looked fine. But they have no usable data about their customer, no creative intelligence, and no ability to move budget when platform costs spike.
A Smarter Approach to AI Ad Spend Across the Walled Gardens
Don’t abandon Google, Meta, or Amazon. For most CPG and DTC brands, those platforms still deliver reach and intent signals that matter. The answer is to enter them deliberately, with structure.
Here is how we think about it:
- Set campaign objectives before the algorithm does. Define what a conversion means to your brand, not just to the platform. If you’re running Performance Max, use asset group segmentation and audience signals to constrain the automation, not just fuel it.
- Bring your own first-party data. Customer lists, CRM segments, purchase history data. These inputs reshape how the AI targets and reduce the platform’s ability to use broad prospecting that serves its yield, not your goals. We have covered why first-party data wins in AI-powered programmatic environments.
- Preserve budget outside the walled gardens. Retail media networks beyond the big three, connected TV, creator-led content, and emerging channels like ChatGPT advertising are all building real audience access. Diversification isn’t just risk management; it’s negotiating leverage.
- Audit your automation quarterly. What percentage of your spend is fully automated with no human override? If it’s above 60%, you’re operating without a steering wheel.
- Tie platform performance to business outcomes, not platform metrics. If your Google AI campaign is hitting a 4x ROAS but your revenue is flat, the algorithm is optimizing for something that isn’t your actual goal.
Human Expertise Is the Unfair Advantage AI Can’t Build for You
Platforms will keep improving their AI, but that isn’t the competition. The competition is whether you have people who can read what the AI is doing, challenge it when it’s wrong, and build strategy across channels that no single platform’s tool can see.
At Junction 37, we believe in human expertise enhanced by AI, not replaced by it. That is the entire reason an independent performance agency exists.
The brands that win this next era of advertising will not be the ones who automate the most. They will be the ones who automate the right things and keep the strategic layer human.
If you’re a CPG or DTC brand that wants a clear-eyed media mix strategy built for this environment, let’s talk.
—
Frequently Asked Questions
What are AI ad spend walled gardens and why do they matter for CPG brands?
AI ad spend walled gardens refer to the closed advertising ecosystems of Google, Meta, and Amazon, where AI-powered tools manage targeting, placement, and bidding internally. For CPG brands, they matter because these systems limit data portability, reduce transparency, and create dependency on a platform’s own optimization logic rather than the brand’s business goals.
Should CPG and DTC brands stop using Google Performance Max or Meta Advantage+?
No. These tools generate real results, especially for conversion-focused campaigns. The issue is running them without structural constraints. Brands should bring first-party audience signals, define campaign objectives clearly, and segment asset groups to limit fully unconstrained automation.
How does AI-driven media buying concentrate power with Big Tech?
AI tools built by Google, Meta, and Amazon are optimized to maximize their own ad inventory value. When brands automate fully, the platform decides placement, audience, and creative weighting. Budget naturally flows toward the platform’s highest-yield inventory, which increases advertiser dependency and over time raises costs for everyone competing in those auctions.
What is a practical first step for a brand that wants to reduce walled garden dependency?
Audit your current media mix and identify what percentage of spend is fully automated with no human review or structural constraints. Then identify one channel outside the big three, whether that is a retail media network, connected TV, or an emerging platform, and allocate a defined testing budget. Real diversification starts with a committed dollar amount, not a stated intention.
—
Chris Pyne, Founder and CEO of Junction 37 and its sister venture Series A. He built Cortex, J37’s proprietary AI-driven planning ecosystem, and pioneered the integration of predictive marketing science into client strategy. Previously, Chris held C-suite roles at OMD USA and MediaCom, where he led planning for $7B in billings and 700+ employees.