Rob Beeler
AI dominates the discourse of our industry and so I’m keen to find examples of companies using AI – especially if it means potential revenue gains or business insights for publishers. Relevent Digital published a blog post: Relevant Digital Launches Relevant AI – An Intelligent Assistant to Revolutionise Ad Monetisation and Analytics and I asked Petri Kokkonen, CEO & Partner at Relevant Digital some follow up questions.
Beeler: Have you had a chance to play with this yourself? As someone like me—who has pulled reports, created pivot tables, and basically become an Excel wizard—how does it feel to just focus on the question at hand?
Kokkonen: Oh yes, I’ve been playing around with our AI quite a bit and have seen it evolve throughout the integration and training process. Things got exciting once we integrated the AI into real data sets.
Most of our customers use Relevant Yield to report historical sales and inventory data (collected through APIs from SSPs and ad servers) as well as real-time revenue data. So, the number of data points can get largish pretty quickly. This is a great example of where AI can add significant value. You can make queries about, say, last week’s performance—pulling from both API-collected data and log-level setup data from your server and client-side setup. The answer could be a combination of insights from both sources, highlighting pros and cons, anomalies, advertiser-level trends, and more.
Of course, you could do all of this manually, but AI saves time and often surfaces insights you wouldn’t immediately spot in manually built reports.
And don’t worry—AI isn’t here to kill the joy of life for true Excel nerds. If tabling data is what tickles you, AI can actually inspire you to build even more complex tables!
On the other hand, not everyone is a pivot table wizard. For those users, AI makes it possible to run sophisticated analyses and keep exploring data—without spending hours wrestling with ChatGPT just to learn Excel.
Beeler: Do you have clients or users actively working with it? Who were the first to adopt it, and what kinds of questions did they tackle?
Kokkonen: Yes, we do, and the initial feedback has been both excited and surprised. Currently, we’re granting access only to admin-level users, and as you can imagine, they immediately start pushing the AI to its limits.
One of the first pieces of feedback we received was from a customer who asked the AI to build year-over-year advertiser-level comparisons. The goal was to easily analyze changes in the advertiser landscape. This is a great example of a real-life use case—AI quickly identifies which advertisers are missing, who has increased their investments, and how key metrics like eCPM have evolved in both direct deals and the open marketplace at an advertiser level.
Our customers also see the value of AI in supporting sales teams, and we’re planning to expand access to all user levels soon.
Beeler: I expect we’ll see many adtech solutions launch with AI capabilities—what sets yours apart?
Kokkonen: AI is being integrated everywhere, and it’s likely that all adtech companies will introduce something in this space—just as we’re seeing across other SaaS verticals.
We’ve spent a lot of time ensuring our solution delivers real value, rather than just slapping an “AI” sticker on an existing product. And that’s the real question: does AI provide meaningful value for the end user, or is it just a shiny toy you play with for a while and then forget?
Our approach to AI value creation is that it’s deeply integrated into each customer’s unique dataset. It’s their data that AI processes, meaning the insights and findings are unique to them. We’ve also integrated internal and industry-specific technical knowledge, making it easier than ever to get guidance on setup-related technical issues.
Maybe you’ve forgotten how to add a new server-side bidder, configure a UID provider, or interpret an especially nerdy metric? AI has your back. And some users might even find it helpful to get assistance in their native language.
Unlike your team’s tech or Excel wizard, AI never gets tired of your questions. If the first answer isn’t clear enough—just keep pushing forward!