As I begin to write, I realize the art of writing blog posts is slowly becoming as much of an “archaic art form” as writing in cursive. Even several of the topics below were spurred by insights and ideas generated directly from conversational AI.
The age of AI is now, and the speed of its growth, use, and adaptation are at levels that will have a fundamental impact on not only the digital landscape but society as a whole.
Suggestive text has evolved under the lens of Large Language Models (LLM)
Starting as early as 2011, we watched the infancy and beginnings of language models with suggestive text. Simple sentences and statements scoured from websites were offered as headlines and descriptions for ads in an effort to allow advertisers to deploy ads more rapidly.
This quickly evolved into Dynamic search ads where the user was provided, in real-time, with vocabulary from a site, but tailored to the search query and product or service landing page. This effort increased with responsive search ads. Which allowed the AI to analyze which verbiage was interacted with most frequently and which verbiage produced the best lead and sales outcomes.
In 2023, Chat GPT rapidly grew its user base to 13 million unique daily visitors.
Google’s Marketing Live event reintroduced LamDA, its "Language Model for Dialogue Applications" under the branding of Bard, a conversational AI tool.
The difference between the two major tools is in Bard’s ability to respond to open-ended dialog and access to more recent events, but this has also led Bard to be more prone to “AI hallucinations” making up information rather than sourcing it as ChatGPT does.
Because language is nuanced; literal and figurative, informational and emotional, that versatility has always made language one of humanity’s greatest tools and one of computer sciences' greatest puzzles to solve.
Chatbots of the past required rigid structures and narrow pre-defined paths. This prevented the open-ended nature of the conversation. LLMs solve these issues with an ability to engage in a free-flowing way, with access to massive amounts of information and trained in dialog.
LLMs and the Digital landscape
What does this evolution of language models mean for advertisers and agencies?
Faster Deployment Options for Ads
The truth is we are still in the early stages of conversational AI and its impact. At the moment the largest impact within the ad space is the ability to rapidly research new keywords, and quickly deploy multiple iterations of high-quality ad copy.
Advanced Audience signals
With conversational AI being able to understand your search and browsing history, combined with GPS signals and actual vocal conversation, the quality of audience signals is drastically improving in exponential ways with thousands of layered minute signals.
Personalized Targeting
Suddenly an open-ended statement of, “I just learned to play guitar” becomes an opportunity to introduce the music shop down the road or an introduction to lessons.
Yes, your phone has been listening to you, but your audio, TV, and connected devices are also sending signals. Those details can go as far as the grocery list created and saved from your smart fridge.
Platform specific targeting and audiences and the rise of 1st party data
The alternate downside to this has been seen with privacy law. IOS 14 updates along with GDPR and CCPA regulations working to limit these signals. Conversely, user service agreements allow the same data to be housed inside walled gardens that keep your data siloed within a single platform.
Owning and managing your company's first party data is more crucial than ever. It allows you to remain compliant while ensuring your data and your customers are influencing AI decisions and personalization, rather than broad platform audiences and outside cohorts of users that may be brand loyal to competition.
The rise of new measurement under a click-less future
We have become accustomed to the granularity of data that has shaped the last decade of digital advertising, and that's about to change.
I'm certain that reach, frequency, impressions, clicks and ultimately return on ad spend will remain key metrics in advertising. But they cannot and should not remain the only way we measure channel or platform success.
The clickless future of advertising is here, right now.
Think of the last time you traveled and were missing something, you probably had a conversation, did a search, and found what you needed. Walked into the store or called and made a purchase.
Human intervention is still needed
Already we use tools to track footfall, customer calls and several other actions that lead to a purchase that are not a part of an online or digital experience - but were influenced by ad exposure.
Point of sale data and true business analytics that measure data-driven impact are elements we need to consider and accept when evaluating the digital experience to make the best decisions on investment and channel mix. While clicks and return on ad spend are still vital elements - they now require a larger picture view that can often be a difficult web to untangle. Being able to understand the influence of your marketing efforts vs. the return is the new lens we need to consider.
For more information related to AI and its impact on marketing check out the Acadia blog. You can also follow Jeff Cline on Linkedin.
Jeff Cline is a Paid Media Manager at Acadia.