Meta announced it will pass on taxes to advertisers, and the market made a fuss. That's normal. Every time a giant makes a slight change, the tide stirs. But, after the initial shock, a less comfortable question remains: why do we remain so dependent on a few platforms to the point where any adjustment becomes a drama?
The problem isn't the rate. It's monoculture. When you plant everything in the same field, any pest will ruin the crop. It's the same in media: a new policy, a more "temperamental" algorithm, a cost increase, a change in attribution, the end of cookies in Chrome. None of this is new. History is cyclical. The label of the problem changes, but the root remains.
I witnessed this firsthand with a mobility startup. Rapid growth, geographic expansion, that great feeling of having found the right path. At a certain point, the company adopted an AI solution to automate campaigns. It worked so well that they decided to concentrate everything on a single channel and invest 100% in that format. Then came the day when performance plummeted out of nowhere. No configuration changes and no explanation from the system. Since the entire operation was in the hands of the algorithm, there was no black box to open. The model delivered the finished product, but not the recipe, and the result? A scramble to rebuild campaigns, loss of revenue and traction, including team cuts. At the time, they blamed the channel. The mistake wasn't "where" they advertised, but rather depending too much on a single place.
Agencies and advertisers know this truth. They talk about diversification in presentations, but in day-to-day operations, the pressure to meet targets and the temptation of convenience push everything towards the same two or three walled gardens. Meanwhile, movements like Meta serve as a warning: whoever controls the infrastructure dictates the rules. They pursue profitability, like any serious business. They are more than right, and the question is what we do with this warning.
Diversification is not a fad, but a matter of governance. It's about treating media like a financial portfolio, seeking low correlation, balancing risk and return, and ensuring strategic liquidity. When the budget is intelligently spread, a bad tide doesn't become a shipwreck. When it's concentrated, any wave becomes a ripple.
"Okay, but diversify to where?" There are solid paths that, combined, already account for a significant slice of the digital pie in mature markets. Programmatic with quality inventory and clean data. Native advertising that respects context and delivers real-world engagement. Rich media that plays with interaction and recall. In-app media with efficient reach and frequency. Audio that builds brand while keeping up with daily life. Video in premium formats, from CTV to well-positioned mid-roll. It's not about replacing one dependency with another, but about assembling a basket with different roles, clear metrics, and growth hypotheses.
This is where the role of each side comes in. Agencies need to resist the autopilot that prioritizes what is easy to operate and difficult to justify when it goes wrong, and on the advertisers' side, the invitation is to give media buyers the freedom to not just focus on direct responses, and to have room for long-term metrics.
First, an honest diagnosis of the current risk. How much of your CAC depends on Meta and Google combined? If the answer is: "it exceeds 80%", you already know where the danger lies. Then, a period of disciplined exploration. Establish a fund of experiments per quarter, with explicit hypotheses, cost and quality benchmarks, and evaluation windows that respect your business cycle. It's not about playing around with testing. It's about learning methodically. Finally, governance of learning. Every week an insight becomes a course correction. When something performs, don't "fall in love": understand why, document it, replicate it, and define the saturation point before you get there. Media is the mixture of art and science.
Let's return to the startup example. If the media plan had been a portfolio, the sudden drop in the dominant channel would have hurt less and taught more. With diversification, you keep your pulse. Without it, you're stuck at the mercy of systems that don't owe you any explanation.
The discussion about passed-through taxes, rising CPMs, and disappearing attribution signals is valid. It shows the reality of a market seeking profitability and privacy. But using this noise only to complain is to miss the chance to emerge stronger. What matters is how each advertiser and each agency will redesign their own mix so that the next rule change is a sail adjustment, not a shipwreck.
Ultimately, the challenge is less romantic and more operational. How is your plan today? Is it truly diversified, or are you still ignoring the ideal world? Because the ideal world doesn't exist. What exists is the plan you take off the paper, revise, measure, and improve. The question that applies to 2026 – and to any cycle – is just one: do you want to play the platform game as a hostage to its rules, or do you want to take advantage of its incredible resources to build a winning and solid strategy?
By Bruno Oliveira, COO of ADSPLAY
