AI Optimisation is The New Media Buying Frontier

What to Read Next

1. Teads Partners with XPLN.AI to Advance Attention Measurement

2. Factory Berlin, All Voice AI Launch Real-Time Voice Monetisation Platform

3. Continua Adds Amazon, Google, Instagram, & TikTok Shopping to Social AI

4. Airtory, CultureSync Media Partner to Deliver Rich Media Advertising

5. VuePlanner Integrates Sundogs’ Creative Insights into CreativeVue

Several years into a fundamentally changed media-buying landscape, many people still don’t understand what’s now driving digital marketing.

Media buyers are not only living in a post-cookie world but an increasingly AI-dominated one. Together, these two paradigm shifts are responsible for one of the most significant evolutions of digital ad buying of the last twenty years. 

Methods that were once commonplace and fully reliable now fall well short of providing maximum value. Relying solely on Facebook Pixel, for instance, to track user behaviour on ads is now insufficient. The game has changed.

Alongside this shift, AI optimisation (AIO) has rapidly evolved from something that’s nice to have to an essential need. Meanwhile, Conversion API (CAPI) remains the foundation of proper media buying in 2025, and running paid ads without its implementation is tantamount to throwing money away. 

CAPI forms the basis of a strategic, data-driven approach in the post-cookie landscape that will make or break paid media campaigns, and it’s where AIO happens.

In essence, AIO’s goals are much the same as SEO, in that the point is to optimise content for digital engines. Whereas SEO was aimed at getting content to rank high on search engines, AIO is focused on training and platform algorithms to find the right customers and maximise campaign performance. 

AI Optimisation Has Become ‘The Essential’

For media buyers, AIO is not simply about automation. It gives media buyers the ability to strategically feed AI the right data to significantly improve campaign outcomes.

All of this is downstream from a major series of changes on the platforms themselves. On Meta, first came Facebook Pixel to track page views, button clicks, and other browser-side events. Conversion API was developed later, and aligned with the blocking of third-party cookies and the transition toward first-party data. 

In fact, Conversion API was initially rolled out as a mitigation strategy for IOS 14.5, and only later evolved into the standard performance driver. Conversion API was envisioned as an add-on as opposed to a replacement for Pixel, and proper optimisation of media buying campaigns requires both to be used in concert. 

With first-party data now reigning supreme, Conversion API becomes the main tool for AIO. Marketers feed first-party data directly to ad platform algorithms, not solely for the sake of reporting but to train the AI to understand who your customers are. 

Once the AI is trained on the specifics of target customer demographics, it can find more of them. From there, a campaign can explode.

Miscommunication and misrepresentations

To some extent, it has proven difficult for the big social media giants like Meta to effectively communicate the essential nature of the Conversion API to the millions of digital marketers worldwide who use it. AIO is even more frustratingly opaque. 

The goal is not to get digital marketers to abandon Pixel (which they know and like) and replace it with something new. The message is to use it in combination with the Conversion API for AIO. 

The lack of understanding we see on the marketer side is only multiplied by Conversion API’s perceived technical complexity. The tool can feel frustratingly complex, and its benefits are not always immediately obvious. 

Marketers see their Conversion API score and figure it’s good enough, not realising how major an improvement of just 1.5 points could be.

In other words, marketers think they’re score is fine and that AIO is superfluous, and then they don’t feel any urgency to fine-tune and improve it since it’s already technically working. 

The data flows, the ads run, and everyone’s happy—and no one realises they’re missing out on an enormous performance improvement with just a few small adjustments.

Pooling the data and turning on the jets

Activating Conversion API isn’t like turning on the tap, exactly. It’s the faucet itself, and only a powerful and useful one if it has enough of a reservoir of data to pull on. Let’s say your Meta attribution shows 50 conversions out of 100 total purchases. 

If you send all the conversion data to Meta, instead of just the 50 that are attributed directly to it, you double the information that Meta can use to train its algorithm and optimise delivery. You double the size of your reservoir.

But digital marketers are, by and large, only sending the data related to the 50 conversions that Meta made. This is a misunderstanding of what the Conversion API does, and it’s a stumbling block of enormous proportions when blown out to the major social platforms. The bigger the blind spot, the bigger the stumbling block. 

Conversion API works extremely well, and it works brilliantly for AIO. The struggle to get ad buyers to use it properly isn’t just bad for individual campaigns but for the industry at large. Remedying this problem will make ads better, data reservoirs better, and the industry better. There’s virtually no reason not to.

Human media buyers vs. AIO 

The goal of AIO is not to replace media buyers, nor is it designed to do so.

AIO will supercharge media buying campaigns, gleaning the most accurate insights from data and managing repetitive, low-level tasks. High-level strategy, creative decisions, and big-picture choices about direction and goals remain human-led. 

And AIO only works well if the data setup has been meaningfully audited by intelligent marketers.

Moreover, AIO requires continuous human refinement. It is not a set-and-forget solution but an ongoing process subject to continuous testing, editing, and adaptation. The data being fed to the AI must be clean and comprehensive, which is a human-led task.

AI optimisation is fast becoming the new standard for media buying. We already know it works smarter than what came before it, and that it’s simply essential in a landscape of LLMs and AI-powered search. Media buyers can only stay competitive with a strategic, AI-powered approach. 

In a market as challenging as this, treating it as less than essential will only result in ad buyers and their campaigns being left behind.

ALSO READ: How Can Brands Escape the AI-Sameness Trap?