Using AI for Agency Success: Balancing AI and Data

By : Robert Wollner


The rise of AI in advertising is like navigating uncharted territories in the Wild West. There are promises of a better, easier life for everyone, with remarkable potential!  AI will automate processes, optimize campaigns, place media, and even extract valuable insights about creative assets. But you need a plan to navigate this Oregon Trail or dare you risk your oxen’s health and brand integrity! One has to have a pioneering attitude, but also plan to not rely too heavily on AI-generated content without proper oversight, which can lead to misinformed ads and diminishing returns. Agencies should approach the challenge of AI-driven advertising with caution, focusing more on how to integrate data into the AI process instead of relying on AI to be  the sole driver of content.

Gen AI is a Co-Pilot

Our Chief Technology & Product Officer, Mickey Alon, put it best: AI serves as a co-pilot, not a replacement for human creativity and expertise. While generative AI algorithms can process vast amounts of data quickly and generate content according to those inputs, the prompts still need expert context to offer relevant and meaningful content. AI needs to be continually informed and refined by human input, to transform a big idea into resonant derivative creative for various platforms. The synergy between human expertise and AI-driven analysis is what will make modern advertising effective.

That’s why agencies are even more critical than ever to brands. Agencies have a deep understanding of their clients, often better than the clients know themselves. They bring in-depth research, understanding marketing behavior and the competitive landscape, as well as the ability to pair creative with ideal media distribution strategies. Agencies and the brands still need to work together to come up with the original “big idea” before anything is sent to a Gen AI tool. Once the on-brand hero asset has landed, Gen AI works wonderfully for things like brainstorming creative by platform and versioning for multiple platforms and formats.

Understanding the Limits of AI

One of the biggest challenges companies are facing right now is choosing the right Gen AI stack for creative applications from a plethora of available options. How do you know which one is better? Platforms like Gemini, DallE, Firefly, and others offer similar value and services with the ability to generate ads and do it at breakneck speeds. The issue is that without a brand’s creative first party data, these Gen AI outputs will have no brand distinction and will  inevitably mirror your competitors’, lacking the nuanced understanding required to generate brand-specific, effective advertisements. Agencies that rely solely on generic Gen AI solutions risk delivering commoditized results, which could ultimately reflect in lower win rates. A brand’s creative first party data piece is a huge missing piece in a lot of procurement conversations and will result in wasted budget lines. 

The good news is in all of this, AI has unlocked the ability to break down creative assets and understand which creative elements are tied to performance. This AI-driven data is the key component to what can make GenAI tools produce optimal results when paired with (human) experts to provide further context. A real-life example of this was illustrated when Vidmob partnered with a luxury cruise line to analyze their creative ad performance. The data showed that ads featuring horses and waterfalls were performing well against the desired KPIs. Without teams who can analyze creative data to provide relevant context and data, an uninformed AI model might assume that these elements alone drove the success, leading to the creation of ads filled with these visuals that could look like these: (The ads definitely stand out and promote a cruise venturing off waterfalls!).) 

But upon deeper analysis of the data, our strategists realized these elements signaled the excursions and activities beyond the cruise ship, highlighting what travelers were interested in. This contextual understanding, validated by human reinforcement, informed a more effective strategy that could be adapted for different platforms like Meta and TikTok. This example highlights the importance of three elements needed for successful use of AI in advertising: creative data, smart teams (agencies!), and the right AI solution. You need all three to be successful.

Empowering Agencies with Creative Data

It’s this unique creative first party data, based on historical performance, current performance metrics, desired KPIs, platform best practices, and brand guidelines that Vidmob can provide to establish a sustainable, competitive advantage for agencies and their brands. Our proprietary, AI-driven creative data platform processes trillions of data points, creates classifications then distills them into custom insights, tailored specific to a client’s needs. Having this level of creative data plus intelligence to inform marketing and media decisions is crucial for developing successful campaigns and it can’t be achieved with commoditized AI solutions. Creative data is an offer in agency RFPs that will stand out, so make sure to choose a creative data solution one with this type of differentiation in outputs.

By using creative data as a foundation, agencies can craft the best big idea based on client specific best performance, and inform Gen AI prompts for optimal performance. But not only does creative data stand out as a competitive advantage in creative production for clients, it also benefits agencies in allocation of media budgets. Agencies can ensure that all creative assets submitted to platforms meet minimum best practices and brand guidelines, guaranteeing the highest baseline performance against algorithms to maximize media spend. Then with continual learnings on the creative submitted, the production can be informed accordingly to become a flywheel of performant assets.

So my agency friends, venture bravely into the Wild West of AI, where the lands promise relevant, contextually informed, and ultimately effective ads. You’re the brave pioneers leading the charge so saddle up your trusty creative data steeds, grab your resourceful sidekicks, and navigate the frontier with wisdom. Good Luck on this new “Oregon Trail” may you find the AI future you’re looking for…and hopefully avoid dysentery.