Adtech has had a tendency to solve complexity, but create a new set of issues as a result. In this thoughtful guest article, Ian Maxwell, CEO at digital media trading platform Converge explains why AI shouldn’t sit as a layer on-top of this complexity, but provide fresh and innovative solutions that don’t need solving.
The years-long effort to prune the tangled digital advertising supply chain is bearing fruit, and now the branches of DSPs and SSPs are overlapping. As the gap between advertiser, audience, and publisher closes, the quality of behind-the-scenes AI becomes the primary differentiator between buying platforms. With everything from audience selection to real-time optimisation now automatable, the intelligence of the supply chain matters far more than its scale.
This conflation of terms that are becoming increasingly meaningless is an opportunity for a reset. Time to step back and look at the programmatic supply chain and ask, why are we doing this in the first place? Adtech has a long history of causing problems, developing solutions to solve them, and so creating new problems. It’s an industry that spends more time plastering over the cracks than fixing the foundations.
Rather than layer AI on top of this mess, the technology is ideally suited for offering a fresh start. Programmatic trading was created to handle the incomprehensible scale of buying and selling ads on the internet. It’s exactly the kind of complexity that AI models are ideally suited for solving, offering a shortcut through the supply chain’s many twists and turns, and dodging the DSP/SSP distinction entirely.
Collaboration Unlocks the True Strengths of AI Adaptability
There’s been a lot of romanticised man versus machine imagery thrown around in the AI debate which misses the forest for the trees. There isn’t some ultimate algorithm we can develop that will do all our work for us while we sit back and have a cup of tea.
AI models need to be set up and trained for a specific task and then deployed after the fact to automate it. This makes AI most effective when it is developed collaboratively: the agency comes to a vendor with a problem, the vendor goes away to see if AI can solve it, and then (hopefully), voila!
In fact, it is this very adaptability that makes AI so useful. While AI can be deployed in plug-and-play, set-and-forget solutions, it’s bespoke AI that has the most potential to meaningfully streamline agency operations.
A platform that layers a chatbot over its user interface or automates mundane processes can stamp AI on its promotional materials and claim to have transformed marketing forever while the nuts and bolts of the system remain in place. Once AI becomes rudimentary and not the shiny new buzzword it is today, how much will these platforms have really changed?
What is far more meaningful is a platform that allows agencies to get AI working for them, solving their specific challenges wherever the technology’s powers to automate, predict, and calculate at scale might be useful. No two agencies are the same and AI, for the first time, allows solutions to be rapidly developed to match that reality.
Ironically, AI Offers a More Human Way of Trading
It’s worth noting that this talk of platforms masks the reality of what makes AI work best for agencies: people. It’s people who agencies talk to about their challenges, and people who go away and see if AI can solve them. Being able to have a normal conversation is an underappreciated skill in the adtech world, where you’re more likely to be bamboozled by acronyms than simply asked, “What do you need?”
The truth is, agencies don’t care whether they’re buying inventory and audience access on a DSP or an SSP. They don’t care about the nitty gritty of programmatic trading, and why should they? It’s a market at the end of the day, and we shouldn’t waste any time telling buyers what wood our stalls are made of and which drill we used to put them together when all they want to do is make a purchase.
All agencies want are targetable audiences and controls for reach, frequency, and outcomes. Maybe some brand safety in the mix. You don’t need to learn a whole second industry-specific language to communicate that, or to deliver solutions that offer what they need.
What agencies are tired of is feeling like resellers moving cash from A to B, bound to platforms built to control the market more so than they are to facilitate trade. AI offers agencies access to tailor made solutions to their (or their clients’) specific problems, dodging bloated tech stacks and a supply chain that’s still trying to solve problems created by adtech a decade ago.
AI Takes Tech Off the Table and Turns Talk to Outcomes
Between the convergence of SSPs and DSPs, the emergence of powerful AI, and the increasing calls from clients to cut complexities, it’s clear that the adtech sector needs to do some soul searching. The years of self-serving huff, puff, and hyperbole around the technical prowess of various backend solutions have failed to put even the smallest dent in the dominance of walled gardens, and few would look at open web trading today and feel that all the effort has paid off.
Overhauling the supply chain’s technological underpinnings with agentic AI gives ad tech the opportunity to step back and simplify. We must resist the urge to dazzle clients with new high-tech jargon (this never works) and instead bite our tongues and deliver the services they need, in the language that speaks to them.
The advertisers who fund this whole circus want to speak the language of business. They want to be able to track the return on their marketing investment, their increased net profit, and the uplift of brand awareness over time. Agentic AI serves as an interface between these simple, top-level goals and the wealth of ad space that can deliver them, and for once it’s advertising technology that can disappear into the background and simply do its job.
What advertisers want less of, frankly, is waffle from the adtech ‘chatterati’ and the overwhelming, tortuous language that’s attached itself to putting ads on the internet.
It’s not about SSPs and DSPs at the end of the day. It’s about connecting supply and demand, and whichever platform can do that for agencies is what they’ll use. AI solutions built from the ground up for such a purpose are going to leave cumbersome legacy solutions in the dust.
Successfully delivering such solutions is more down to having integrated human-to-human conversations with one another. Not having bolted-on technology that could soon be table stakes.
