AI has the potential to deliver unparalleled data-driven insights that transform marketing strategies. But achieving this potential requires more than just adoption—it demands a disciplined, strategic approach.
Technology
Services
Resources
Amidst all the networking, socializing, and poolside festivities at this year's Possible in Miami Beach, there was one topic that dominated all the rest: AI.
Throughout the event, artificial intelligence was on the top of attendees' minds and the tips of the speakers' tongues. How are you using it? Where can it provide the most (and least) benefits? What are the keys to harnessing AI’s power successfully? And how can business leaders fully exploit the potential of AI at their organizations while mitigating its safety risks?
The clearest takeaway from the industry's thought leaders? Marketing's AI revolution has officially commenced.
Here are some of the top insights to come out of the Fontainebleau in Miami Beach at this year's event.
Early usage indicates that AI could lead to a productivity boom, allowing marketers to get more done in less time. A Microsoft study found that users of its Copilot AI tool spend less time writing emails, summarizing long documents, and completing first drafts of documents. AI can “attend” meetings for us and summarize the key findings and takeaways. It can also provide new context around how those meetings are conducted, combing through transcripts to identify blind spots in the conversation and capture the tone and sentiment to help people improve the way they show up to work. When used all together, AI can help make us not just more productive, but better marketers and professionals.
In just the last few years, AI-powered capabilities grown at an extraordinary rate, and the idea of AI disrupting every aspect of our lives appears to be on the horizon—including marketing.
Rex Briggs, Chief AI Officer at Claritas and author of the forthcoming book “The AI Conundrum,” noted that marketing is, in many ways, an “ideal use case” for AI.
With AI, marketers can now optimize campaigns in real time and with greater precision, leveraging AI’s proficiency at recognizing patterns and creating hyper-nuanced segments dynamically and on-the-fly to drive desired outcomes. The individual results of these optimizations can be small, but when added up, they can result in huge gains over baseline performance. What’s still to come (but appears to be just on the horizon) is enhancing the technology so that it can get better at explaining precisely what—and why—the AI is optimizing ads based on various criteria for more real-time transparency and clarity.
However, despite all the hubbub around technological innovation, we mere mortals still have a very valuable and unique role to play. Human marketers will be essential in making the final decisions around what to use—and what to change—throughout the campaign process. To skilled marketers, AI represents a powerful new tool to help them move faster, be more productive, and grow more effective.
Google re-iterated its commitment to both the Privacy Sandbox and to deprecating third-party cookies in Chrome by the end of 2024. But Amit Varia, Director of Google’s Privacy Sandbox, emphasized that Privacy Sandbox APIs are not intended to be a 1:1 “replacement” for third-party cookies, and that marketers are best suited to leveraging a range of identity solutions in unison as part of a larger privacy-friendly toolkit. And while initial users of these Privacy Sandbox APIs are seeing effective results, we are still effectively operating in a test environment, where just 1% of Chrome users are playing in the Privacy Sandbox sans third-party cookies, and even Privacy Sandbox evangelists noted there are still ample questions about how these solutions will perform at scale.
So, what value can AI provide marketers in a cookieless environment? MMA data showed that AI-powered personalization drove a 35-65% increase in ad performance within contextual environments. And AI can optimize first-party data to better target existing audience and to create new lookalike audiences.
Audio and display ads both play to generative AI’s current strengths, making them ideal channels for marketers looking to experiment with AI-generated assets and campaigns.
Progressive, for instance, has begun using AI to create more personalized audio ads. The insurance giant incorporated AI across every step of the campaign process, allowing them to go from brief to approval in just 6 weeks (vs. the 22 weeks it previously took them without AI). Leveraging AI-generated scripts and voice talent—after training the AI on Progressive’s brand and an extensive content archive—Progressive’s team was able to create 96 ads in a single week, then run and test them using dynamic creative optimization (DCO) to progressively adapt the ads as market conditions changed. In doing so, they were able to assess different ad parts and predict the right combination for the right audience.
“You can train AI on your own content, hit a button, and end up with different scripts, and then when you are happy with the scripts, you can hit another button to develop the audio, and then when you are happy with those, you can move on to approvals,” said Remi Kent, CMO at Progressive.
Though AI generated the scripts, the personas, and the background music ads, humans were involved and instrumental in every step of the process, in what Kent described as a “collaboration” between humans and generative AI. The result was a process that allowed Progressive to move faster and create more ads with more personalization—all at scale. By using AI-generated ads and leveraging AI-powered optimizations, Progressive was able to drive a 197% lift in quotes over baseline.
Effective utilization of AI for targeting, attribution, and optimization relies on high-quality (and high quantities) of data. The problem? Silos. So very many silos. Disconnected channels, siloed platforms, walled gardens, and a general lack of transparency and data centralization is all too often resulting in organizations having an incomplete picture of their data.
In a session on how marketers can “hack their adtech,” Richard Brandolino, Global Media Channels & Adtech Leader at IBM, recommended that agencies and brands keen on exploring (and exploiting) AI’s benefits should look to facilitate cohesion and interconnectivity across their tech stacks. Streamlining and unifying data flows can yield significant improvements in profitability, cost efficiency, and strategic agility.
Beyond even efficiency and improved campaign results, Possible speakers addressed one other aspect of AI in marketing: the balance between AI’s promise of innovation vs. its inherent risks.
Jaime Teevan, Chief Scientist at Microsoft, noted that the decisions we make with AI today will influence the future of jobs, our industry, and our world. Since introducing its AI-powered Copilot last year, Microsoft has strong feedback that the technology is making people more efficient and saving them time. What’s left to determine is what will people do with that time.
At its best, AI has the potential to create a generational opportunity for innovation, but marketers’ experimentation should always be accompanied by careful consideration about not just the immediate impact of those decisions, but the third- (and fourth- and fifth-) degree effects of those decisions.
Teevan noted that, when first beginning to explore how Microsoft could incorporate OpenAI’s GPT 4 into its products, she began from a place of “How do we bring this technology to people, and do so in a responsible way?” Perhaps tellingly, Teevan’s implication throughout a session on AI seemed to be that the “responsible way” Microsoft has landed upon is to outsource much of this responsibility to its users, imploring them to do their own research, experimentation, and exploration with the technology and hoping they do so responsibly.
Curious about how leading marketers are using generative AI? Basis surveyed over 200 marketing and advertising professionals from top agencies and brands, brands, and publishers to see how marketers are feeling about AI today and gauge how they think it will shape the industry going forward.