Will AI replace media buyers? The honest answer for 2026

The job that disappeared

Go back ten years and paid media was a craft of granular control. Audience segmentation was a competitive skill. You layered demographics, interests, behaviours, and lookalikes. You built dozens of hyper-targeted audience pools. You tweaked bids at the ad level – in some setups, every fifteen minutes across thousands of campaigns. The people who were great at this were genuinely great. It was precise, technical work.

Then, gradually, the platforms took it away.

Interest-based audiences gave way to nested lookalikes. Nested lookalikes gave way to broad targeting. Manual bidding gave way to automated bidding. Thousands of campaigns collapsed into four or five. Each step handed another lever back to the algorithm.

As Nest’s Head of Social, Josh Goldsmith puts it, having spent fifteen years in the space from the early days of Facebook advertising: “Who’s doing bidding these days? Automatic bidding’s come in and that’s the norm now.”

Audience targeting, bid management, budget allocation at the ad set level – these were the core skills of paid media a decade ago. They’re now largely automated. That’s not speculation. It’s the operating reality of every major ad platform.

What actually happened

This isn’t a gradual decline. It’s a structural shift in what platforms need from advertisers.

Meta, Google, and TikTok have spent billions building AI systems that match ads to users. Meta’s algorithm can now evaluate tens of millions of ads per impression – orders of magnitude more than the few thousand it processed before. Google’s Performance Max automates bidding and placement across its entire inventory.

These systems are extraordinarily good at predicting which user will respond to which ad. Better than any human media buyer, regardless of experience.

But they can only choose from what you give them.

If you give the system twenty ads, it has twenty options. If a competitor gives it five hundred genuinely different ads, they have five hundred options. The algorithm rewards the advertiser with more creative supply – more opportunities to find the right match for each user.

This is the shift that matters. Performance is no longer dictated by how clever you are at platform optimisation. It’s dictated by creative supply. The brands that recognised this early and scaled their creative volume saw the results. At Nest, clients running significantly more creative volume through Q4 2025 saw +18% ROAS and +38% revenue year-on-year – and the primary variable that changed was creative volume, running at roughly 3x the previous year’s output.

Even the analytical work that replaced tactical execution is being absorbed. In February 2026, Meta embedded Manus AI directly into Ads Manager – an autonomous agent handling report building, audience research, and campaign analysis. Another layer of the job, moved inside the platform.

The ceiling moved. It used to sit over media execution. Now it sits over creative production.

The role that replaced it

So what’s left for media buyers? More than you might think – but it’s a different job.

Setting the guardrails

The algorithm does the work now. But someone has to tell it what to optimise for. That means aligning your account structure with your business objectives, your unit economics, your growth targets, and your customer personas. Get those inputs wrong and the system will optimise efficiently toward the wrong outcome. The media buyers who matter now are the ones who understand the business well enough to set those guardrails correctly. Not which bid strategy to pick, but what the business actually needs the system to achieve and how to structure the account so it can learn.

Reading the data and knowing what to make next

This is where the role gets genuinely interesting. Performance data now tells you more about creative than it does about audiences. Which hooks are landing. Which formats work for which types of customer. The best paid media people are spending their time here – interpreting what the algorithm is doing with the creative it has, and translating that into direction for the next round of production. Not building audiences. Not tweaking bids. Reading the signal and acting on it, fast enough that the insight is still relevant when the new ads go live.

Structuring accounts for the algorithm to learn

The old version of account structure was about control – thousands of campaigns, each one a lever you could pull. The new version is about giving the algorithm room to learn. That means organising around customer segments rather than granular targeting, running enough volume within each segment for the system to find patterns, and building a testing framework that tells you what’s working at the segment level rather than the ad level. Josh describes the shift simply: you used to have thousands of budgets to manage daily. Now you might have three or four. The complexity moved from the account to the creative.

Connecting teams that used to work in silos

The performance team knows what’s working. The creative team makes the ads. The data team analyses results. In most organisations, those three groups sit apart – and the gap between them is where performance gets lost. Insight reaches the creative team late. Production runs on a cycle that doesn’t match the speed the platforms demand. The media buyer who can sit in the middle of that – translating data into briefs, getting production moving quickly, keeping the whole thing aligned – is doing the job that actually determines performance now.

The honest answer

Will AI replace media buyers?

It already replaced the old version of the job. If you define media buying as audience building, bid management, and tactical campaign optimisation, that work is largely done by machines now.

But it created a different job in its place. One that sits at the intersection of performance data and creative production. One that requires business judgment, not just platform knowledge. One that’s harder, more strategic, and – for people who make the shift – more interesting.

The question isn’t whether AI will replace you. It’s whether you’ve redirected your attention to the thing that actually determines performance now: creative supply, creative diversity, and the speed at which insights turn into new ads.

That’s the bottleneck. That’s where the work is.

How Hummingbird fits

If the role of a media buyer shifts from optimising campaigns to directing creative supply, the constraint becomes production capacity. You can read the data. You know what creative the algorithm needs. But getting it made – at the volume and speed platforms now demand – breaks traditional production workflows.

Hummingbird is the system Nest built to remove that constraint. It connects performance data directly to creative production, so the insights media teams surface on Tuesday can ship as new creative by Thursday. Not next quarter. Not after the next brand shoot. This week.

That’s not about replacing media buyer judgment. It’s about removing the bottleneck that sits between their judgment and the ads that reach the platform.

And there’s a second constraint most brands haven’t clocked yet: strategic alignment.

Early signals suggest Manus is strong on data analysis but weak on campaign strategy. The recommendations it generates look a lot like what a Meta rep would tell you – which means they’re structurally aligned with Meta’s incentive: spend more. They’re not linked to your margin targets, your unit economics, or your growth model. That’s not a criticism of the technology. It’s built to serve millions of advertisers. It’s not built to maximise performance for your brand specifically.

This is why we built Hummingbird the way we did – performance experts training the technology to drive results for scaled ecommerce and retail brands. The AI handles the heavy lifting on data, but human expertise ensures every decision is tied to actual business goals, not just platform recommendations.