As the whole world goes bonkers for generative-AI, advertising agencies are trying real hard to impress their clients and investors with their mad skills in AI and their tight connections to the big players like Google, Nvidia, OpenAI, and Microsoft.
These agencies, while advising the biggest companies on adopting AI, are starting to realize that they themselves are getting disrupted by this technology. They gotta figure out what to automate, where to find AI talent, and how to balance the potential boost in revenue with the risks involved.
Now, some of the challenges they face are pretty standard stuff, like avoiding copyright issues and being transparent about AI-generated content. But guess what? The ad industry is encountering a whole bunch of new problems they never saw coming.
Check it out, some of these agencies are worried that there won’t be enough graphic processing units (GPUs) to go around, and that’s gonna mess with their client work. You see, with the rise of AI, GPUs are in super high demand. These chips are like the superheroes that can process huge amounts of data and media at lightning speed. So now, companies are being divided into the “GPU haves and have-nots.”
Right now, most agencies are renting compute power from cloud service providers like Amazon’s AWS, Microsoft’s Azure, and Google Cloud. It’s all good for now because these tech giants and well-funded startups are competing like crazy, which keeps the prices relatively low. But as the market matures, agencies expect the prices to go up. And here’s the kicker: agencies are promising their clients that they can churn out thousands of awesome creative assets using generative-AI. So they’ll eventually need to get their own compute power.
If they don’t plan ahead, they might end up at the back of the line, waiting for months to get their hands on those precious GPUs. Lewis Smithingham, innovation expert at Media.Monks, says agencies might either get charged way too much or fail to deliver on their promises because their in-house compute is outdated or they underestimated the amount of content they need to produce.
Now, here’s another thing. AI is making agencies more productive and efficient, which sounds great, but it also means that agencies might have to lower their fees. You see, they usually charge based on the number of hours, like full-time employee hours, that they spend on a client’s business. But with AI doing a lot of the heavy lifting, it questions whether that traditional compensation model is sustainable.
According to Ryan Kangisser from the consultancy MediaSense, generative AI makes them question how they can keep the lights on with time and material-based compensation. Some experts suggest that agencies might need to switch from FTE to HTE – “human plus technology equivalents.” The work being delivered is not just human hours but also AI, data, and expertise, says Jay Pattisal from Forrester.
And as it turns out, agencies don’t really know what to do with all the AI talent they’ve hired. They’re like, “Cool, we got the AI people, but where do we put them?” Most of them try to fit AI under the chief technology officer or the IT department. But Ruben Schreurs from Ebiquity says that AI goes way beyond just technology. It has broad business impacts and comes with legal and compliance risks. So agencies need someone in charge who can make the necessary decisions and recommendations about agency structure, culture, and purpose.
It’s a wild ride for the ad industry, trying to navigate the AI wave. But hey, they’re adapting and figuring things out so they can stay ahead of the game.