Sam Scott returns to mark the homework of 2024 marketing predictions in his column, The Promotion Fix. It’s an annual tradition that holds charlatans and the misguided to account. This year, his eye turns to AI.
While AI hype is meant to impress investors, analysts, and journalists, consumers actually do not care whether a product uses a glorified computer program. The research shows us this time and time again.
Research published this month in Harvard Business Review stated that people do not want AI that seems too human. Furthermore, consumers are reportedly distrustful of advertising that bombards them with mentions of the technology. With that in mind, I looked at predictions made about 2024, specifically about AI in marketing, scoring each for accuracy.
In 2018, I looked at Deloitte, Forrester, Gartner, Scott Galloway, HubSpot and Salesforce. (Here are the results.) The following year, GroupM, Kantar, MoffettNathanson Research and the World Federation of Advertisers went under my microscope. (See who won.) In 2021, it was Future Today Institute, GWI (formerly GlobalWebIndex), McKinsey, and Zenith Media’s turn. (Look at the scores.) Then, I looked back at many of the predictions that were made as the coronavirus pandemic was starting. (The results might be surprising.)
It’s cathartic to separate the wheat from the chaff. And it’s particularly important at a time when AI has the power to change entire industries.
So, what is AI?
In 2024, I became addicted to Out of the Park Baseball 25 (OOTP). The simulation video game focuses on the 90% of the sport that occurs outside of the gameplay on the field – things such as statistical analysis, minor league player development, front office staff, trading, lineup creation, and revenue from ticket sales.
It uses “artificial intelligence” to run simulations of games or even entire seasons – just like how marketers can use AI to simulate the writing of copy and the creation of advertisements. It uses “AI” as a buzzword for computer programs that simulate human behavior by using training data and predictive analytics to make decisions or create some type of output.
Want to go deeper? Ask The Drum
In OOTP, the game is trained on real, historical baseball data that it then uses to make decisions on likely outcomes. Say a pitcher is about to throw a pitch. The program seemingly uses all the training information – from the state of the game to their historical results to determine the probabilities that he would choose various types of pitches.
The program then seemingly does the same thing to determine the batter’s likely action. Every matchup comes with a data set with potential outcomes and varying likelihoods, and then the computer rolls the dice.
This process repeats. And repeats. And repeats
A similar thing occurs when marketers use AI to create the aforementioned copy and advertising. In a March 2023 column, I showed how ChatGPT and other similar AIs use such a process that essentially makes them gigantic versions of Google autocomplete based on calculating what set of words would likely follow a given set of prior words.
So now, for 2025, let’s see how those AI predictions panned out. I reviewed the predictions of Ogilvy, IMG, and Publicis Media Content & Innovation (PMCI).
Ogilvy
Ogilvy released its 2024 influencer marketing trends report in November 2023. One important section focused on AI.
“Meta’s AI Personas, introduced in late 2023 and fully deployed in 2024, signals a significant shift from broad-reaching influence to personalized, one-to-one interactions that maintain a sense of authenticity.”
If you say the word “influencer,” many people will think of a Kardashian. But in marketing today, the use of niche influencers is becoming more popular – and those interactions indeed have less reach but are more authentic. The beauty industry is just one example. In my personal experience, I have seen the use of “micro” B2B influencers to be useful as well.
Score: 1/1
“An upcoming feature even allows users to create personalized AI versions of themselves. This advancement is set to democratize the concept of influence, enabling individuals to have a digital presence that can interact and influence within their own networks.”
A perhaps unlikely source confirms this prediction: OnlyFans models. A Wired report last month found that they get so many messages from people that they are now using AI tools to create replicas that can give them a, uh, hand.
Score: 1/1
“Mascots like Duolingo’s TikTok sensation, the Owl, have already demonstrated the potential of [AI-created] characters in driving social engagement. And Ipsos research shows that investing in characters has the greatest impact on enhancing brand salience.”
Indeed, the use of distinctive brand assets (also called brand codes) is critical in creating mental availability.
Score: 1/1
For more information, I will refer you to my 2022 interview with Jenni Romaniuk of the Ehrenberg-Bass Institute (EBI) on that idea, specifically in a B2B context. (I just wonder if EBI will take issue with Duolingo’s Owl because they have a similar mascot.)
Ogilvy’s total score: 3/3
And this is where I must critique another prior column. In 2017, I published a popular one entitled “TV is not dying – it’s lies, damn lies, and bad media statistics.” Well, in 2024, I dumped my satellite TV provider and have subscribed only to YouTube Premium and MLB.tv for baseball ever since.
For example, I now enjoy following YouTube creators and influencers as they watch TV shows, see movies, and listen to music – all from the 1970s to the 1990s – for the (alleged) first time. Take from that what you will.
IMG
Continuing with the aforementioned sports theme, IMG published a deck in December 2023 stating that “2024 will be the year the sports industry starts applying AI in a meaningful way to transform its business.”
But was that true?
“In 2024 sports organizations will have a clearer idea of how to use AI and will do so to cut (unlock efficiencies in their processes).”
The Leicester Tigers, an English rugby team, was just one example of a sports organization doing exactly that.
Score: 1/1
“Early use cases will see sports using AI to improve ticket and retail conversion through better user targeting. Rights holders will seek to reduce churn on streaming and membership services and optimize pricing through much deeper predictive analytics.”
The Dallas Stars ice hockey team generated more than $300,000 in incremental revenue from single-game ticket price adjustments made using an AI model. Score:
Score: 1/1
“They will also train AI models on content performance data to maximize investment in content production, ensuring they create more of what users value and less of what they don’t.”
The BBC used Stats Perform’s OptaAI suite during Euro 2024 to enhance its coverage, aid with storytelling, and provoke discussions.
Score: 1/1
“Where organizations have been capturing user behavior data, such as in-store or in-stadium movement, AI will provide a valuable layer in improving the live experience for fans by identifying bottlenecks and reducing wait times. Teams and venues will leverage AI chatbots to do much of the heavy lifting of their customer service teams.”
The 2024 Olympic Games in Paris indeed used AI to make the entire operation more efficient.
Score: 1/1
“We will continue to face challenges around deepfakes and misinformation. Athletes and their representatives must be increasingly watchful of which artificial versions of themselves are circulating online.”
A December 2024 deal between YouTube and the Creative Artists Agency gave athletes, among others, early access to YouTube’s tools for finding, reporting, and removing AI-generated videos featuring their likenesses and voices.
Score: 1/1
IMG’s total score: 5/5
Publicis Media Content & Innovation (PCMI)
PCMI released its 2024 predictions in November 2023.
“Anticipate an increase in social interactions with AI-driven agents across platforms and tools that enhance creative output and personalization as a key to unlocking loyal consumer engagement.”
In 2024, LinkedIn unveiled a new AI agent and job coaching tool. Many major technology companies including OpenAI, Microsoft, Google, and Salesforce have done something similar.
Score: 1/1
“Several AI wearables have emerged in 2023 and we think will continue to break barriers in 2024, each with unique features and functions.”
Last year indeed saw the emergence of items such as the Rabbit R1 pocket-sized AI assistant and the Humane AI Pin. But 2025 will tell if they will all turn out to be failed gimmicks or something truly innovative.
Score: 1/1
“In the realm of content creation, and as an iterative tool to personalize creative outputs, brands will leverage AI to enhance creative processes and outcomes across all forms of content.”
One Influencer Marketing Hub study in October 2024 found that 69% of marketers had already integrated AI into their general operations. 42% had indeed specifically used gen AI for comms campaigns.
Score: 1/1
“AI-driven NPCs in gaming experiences: AI-powered non-playable characters will enable brands to engage with players in real-time. With a host of talent like Megan Fox or Nicki Minaj already getting their video game debuts this year, we think brands might even sponsor a celebrity or influencer integration into a game as a brand ambassador in 2024.”
First, every entity that any player has ever encountered in any video game ever has always been an “AI” “non-playable character.” Second, there were many celebrity appearances in video games in 2023 – weirdly, I could not find any in 2024. (Let me know if I am mistaken.)
Score: 0/2
“Socializing with AI chatbots: Offering instant responses to customer inquiries, handling routine tasks, and providing personalized recommendations, or even enhancing live streaming e-commerce – AI-enhanced interactions will become a game changer, starting in 2024.”
According to a G2 report in December 2024, 16% of businesses already use chatbots, and 55% plan to adopt them. I just would not call it “socializing” – that sounds like something creepy out of the movie ‘Her.’
Score: 0.5/1
Be careful: sales that come through chatbots are not necessarily incremental. If someone comes to a website already wanting to buy and they just happen to make the purchase through a chatbot, then the chatbot should not get the “credit.” (See my prior column on the dangers in digital attribution, specifically when I was queuing online for Oasis tickets.)
Try to see if any sales from chatbots are incremental – if not, then they might not be helping your website but rather hurting it because of the annoyance factor.
“DE&I initiatives in the AI space will be a massive opportunity for brands to be on the right side of history.”
In 2024, the New York Times reported that the percentage of quarterly earnings calls that mention diversity, equity, and inclusion had fallen for months while job openings for DEI programs at major US tech companies had also been declining and groups such as Women in Tech and Girls Who Code were closing. And Europe was not much better. Just last week, Meta said that it was ending all DEI initiatives.
Of course, this reality versus the expectation says more about the tech industry than PCMI. As I once told an audience during a keynote speech in 2021 on so-called “brand purpose,” focus on what companies actually do rather than what they say in their advertisements and press releases.
Much of the time, statements about DEI and “cause marketing” are just temporary ways to make more money by hopping on a bandwagon rather than doing what is right over the long term. After all, boards of directors will always ask chief executives how they can make more money rather than ponder their “brand purpose.”
Score: 0.5/1
Publicis Media Content & Innovation’s (PCMI) total score: 4/7
How marketers will actually use AI
Machines have always made processes more efficient, but the outputs are not necessarily more effective. One creative genius came up with the Mona Lisa. All a machine can do is quickly create numerous inferior copies. (See the above slide from a talk of mine on AI and marketing.)
Out of all of the areas that are – or at least should be – under marketing, marcom is the discipline that should use AI the least. AI is good for numbers but not creativity; it just makes things average very quickly. We should use AI to do the most boring tasks – not the most creative ones – because creativity is what makes us human. And we’re just better at it.
Just see this finding from the 2024 State of Marketing AI Report by the Marketing AI Institute. Look at the top. The text and advertisements that companies create with AI get all the attention and news coverage today. But that is not how most marketers are actually using AI in the real world.
I predict that the industry will use AI mainly for market research, product usage analysis, dynamic pricing, distribution efficiency, and getting greater value from existing high-tech tools. Basically, marketers will incorporate AI into every marketing activity except creativity and communications.
As author Joanna Maciejewska put it last year: “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.”
Take the Out of the Park baseball simulation video game that I mentioned at the beginning. The “AI” computer program can take existing statistics and then use predictive analytics to play a game or a whole season.
But it can never replicate the happiness in real life with the crack of a bat, the taste of a hot dog, or the Boston Red Sox defeating the New York Yankees.
The Promotion Fix is an exclusive, long-running column for The Drum, contributed by global keynote marketing speaker Samuel Scott.
link
