Digital advertising is evolving fast. With shifting privacy rules, unpredictable user behavior, and rising competition, AI is no longer a nice-to-have, it’s a necessity. Today, it’s less about hype and more about meeting everyday demands for performance, efficiency, and adaptability.
Peter Kireev, Co-founder and Chief Product Officer at Reliz, the European company behind the machine learning-powered Reliz Ad Platform reflects on the rapid evolution of AdTech. With 14 years in the industry, he considers the past five to be the most transformative as AI moves beyond Big Tech to empower lean, performance-driven teams.
But even with AI’s growing influence, Kireev is quick to remind us where the real magic still lies: in human creativity, intuition, and strategy.
How Would You Assess The Impact Of Artificial Intelligence On Adtech Over The Past 3 To 5 Years?
I’ve been working in AdTech for over 14 years, and the last five have been the most exciting of my career. The reason is how quickly and profoundly artificial intelligence has transformed the advertising landscape.
Today, we can target and optimise campaigns on a fundamentally new level; managing bids at incredible speed, generating thousands of dynamic creatives, and running anti-fraud analysis almost in real-time. AI has permeated every layer of the ad industry and has become an essential part of day-to-day operations.
What’s particularly striking is how accessible these tools have become. What was once the exclusive domain of tech giants like Google or Apple is now within reach for even small teams. In the past, such projects were held back by high infrastructure costs and a lack of quality data. Today, cloud services like Amazon allow companies to deploy environments for training AI models with minimal investment.
Moreover, the rapid advancement of big data technologies has significantly reduced the cost of data collection, normalisation and processing while also expanding the pool of skilled professionals. This has made modern AI solutions more accessible and better aligned with specific business needs.
As a result, AI platforms are increasingly becoming niche and finely tuned for specialised use cases. These are no longer one-size-fits-all solutions, they’re precision tools.
Still, I believe that true creativity and strategic thinking remain human territory. Genuine creativity, audience insights, storytelling; these are things AI can’t replicate. The most effective approach today is a blend: combining the speed and scale of AI with the nuance and originality of human thinking. That’s where real, impactful campaigns are born.
What Role Does AI Play In Your Advertising Products – Where Does It Add The Most Value And Where Does It Still Fall Short?
At Reliz Ad Platform, we take a deeply pragmatic approach to integrating AI, one that enhances core performance without replacing human creativity.
We’ve built what we call an in-house AI model factory; a system for rapidly developing and deploying machine learning models tailored to very specific advertising goals. For example, these may include reducing the cost per install, attracting more precisely targeted user segments, or increasing the number of purchases and other valuable user actions.
Depending on the product and strategy, AI makes it possible to improve audience quality without reducing campaign reach and more cost-effectively. These models help the platform optimise campaign performance, dynamically manage bids, and deliver exact contextual targeting, all while staying firmly aligned with privacy-first principles.
But the true power of AI emerges not in isolation but in balance. The creative concept, the positioning, the overall product strategy, people still lead those. That’s where uniqueness comes from. AI can amplify, accelerate, refine but it can’t invent meaning from scratch.
This merging is already delivering results. At Reliz, we are running high-performance campaigns for mobile apps and achieving outstanding outcomes with significantly lower costs.
Recently, our focus has shifted even further towards specialisation. Instead of developing general-purpose models, we are now building AI systems tailored to the specific needs of individual market segments. Whether it is fintech, gaming, or e-commerce, the more tailored the model, the better the results.
For example, a major ride-hailing aggregator requested a retargeting campaign for its users in Germany. We developed a specialised model trained on the company’s data. The campaign was executed at a minimal cost by leveraging AI while strictly adhering to privacy-first principles.
Which Advertising Tasks Does AI Handle Best Today, And Which Still Rely On Human Input Due To Complexity, Creativity, Or Data Gaps?
When it comes to routine and analytical tasks, AI has become an indispensable force in advertising. It’s incredibly effective at things like automated targeting, bid optimisation, large-scale data analysis, and fraud detection. It helps us react to market shifts faster and more precisely.
It’s still amazing that a well-trained AI model can run a long-term advertising campaign entirely on its own, without any human intervention. That was unthinkable just a few years ago.
But everything changes when the conversation shifts to creativity, brand identity, and emotional storytelling.
These are areas where the human element is irreplaceable. Crafting content that resonates, building innovative concepts, and shaping a truly differentiated brand; that takes creativity, intuition, and experience. These are things AI can’t replicate, at least not yet.
The message is clear: let machines do what they do best and let humans focus on what only we can do.
How Are Advertising Platforms Adapting To The Privacy-First Landscape, And How Does AI Help Make Sense Of Limited Signals As Cookies And IDFA Fade Out?
The industry is still recovering from the IDFA apocalypse and the broader cookie clampdown. Digital advertising operated almost entirely on deterministic data for years; precise user identifiers that allowed hyper-targeted campaigns. But in today’s privacy-first environment, platforms have had to rethink their approach from the ground up.
Advertisers are now shifting focus toward first-party data, behavioural signals, and contextual insights. These changes have sparked a wave of experimentation across the ecosystem. There’s an ongoing debate about which probabilistic targeting and attribution models are most effective in this new landscape and we’re very much a part of those conversations.
At the centre of it all is AI.
Artificial intelligence has become a kind of compass in this post-identifier world. It helps us process fragmented signals, detect hidden patterns and optimise campaigns even when user-level data is scarce or unavailable.
For advertisers navigating this new terrain, the message is clear: data quality matters more than data quantity and AI is the key to unlocking value from the remaining signals.
What Skills Are Becoming Essential For Adtech Professionals, And What Should The “Marketer Of The Future” Start Learning Now To Stay Ahead?
With the rise of privacy-first policies and the fallout from IDFA changes, the skill set required to thrive in AdTech is evolving fast. If you want to stay relevant in this industry today, your number one must-have is data fluency. It’s no longer enough to collect numbers, you must understand how to extract meaningful insights from them.
Today’s marketers need a working knowledge of algorithms, machine learning, and the ability to leverage first-party data effectively. These are no longer “nice-to-have” skills, they’re foundational.
But technical knowledge alone isn’t enough. The marketer of the future is a true hybrid. They need to be both a data analyst and a creative strategist, someone who understands how to build emotional resonance with an audience.
As AI takes on more optimisation and signal analysis, human value increasingly comes from asking the right questions, designing original concepts, and building cohesive, strategic narratives that technology alone can’t deliver.
Machines will do the math, but the magic still comes from us.
