AI and Marketing Data

Yum Brands, the owner of KFC, Taco Bell and Pizza Hut, is seeing increased marketing performance thanks to new bets on artificial intelligence (AI), executives said when discussing the company’s Q3 2024 earnings results with analysts. 

“This quarter, we successfully launched personalized AI-driven marketing campaigns that, relative to traditional digital marketing campaigns, generated significant increases in consumer engagement, leading to increased purchases and a reduction in consumer churn,” said Yum CFO Chris Turner on a call with investors.

Turner declined to break out specific metrics stemming from Yum’s AI-powered marketing initiatives but offered additional color into how the fast food giant’s “AI factory” operates. For example, Taco Bell can accrue data from points of sale and digital menu boards in stores, as well as leverage the chain’s popular loyalty program and app. Linking these sources of intel together and enhancing them with AI allows for more personalized promotions and a refined picture of the customer for other marketing purposes. 

Yum has been running AI pilots across its three largest brands in the U.S. and believes these capabilities will be “broadly and easily scalable” while providing an overall stronger return on marketing investment, according to Turner.

“We’ll continue to bring it to life across the brands and across markets as we progress,” Turner said of AI.

Separately, Taco Bell is expanding its AI voice technology at drive-thrus in the U.S., aiming to implement the solution in hundreds of stores by the end of 2024. Consumer response has been “very positive” so far, Turner said. Other restaurant brands have received backlash from relying on automated technology that can offer wonky and imprecise responses. Yum, in total, has over 40 AI-driven projects in the works. 

Yum’s AI bullishness comes as Taco Bell continues to outperform the rest of the company’s portfolio. Same-store sales at the Mexican chain rose 4% year over year in Q3 2024 and digital transactions soared 30%.

Taco Bell in the quarter also made breakfast optional for franchisees, providing it greater flexibility to spend marketing dollars on growth drivers like its Cantina Chicken and value menus, CEO David Gibbs said on the earnings call.

Building relationships between marketing and data security teams is a growing priority for CMOs, but one-third of those partnerships struggle with collaboration, according to an August 2024 report from the CMO Council and KPMG.  

Among surveyed marketing chiefs, 79% view marketing-security partnerships as important to securing customer data that provides a competitive edge. However, those relationships can be underutilized, with nearly one-third of teams only communicating during times of crisis.  

CMO anxiety around data security is on the rise as the role takes on more functions related to first-party data, customer relationship management and artificial intelligence (AI). Notably, 84% of marketing chiefs believe AI and machine learning could impact data security. 

Marketers at consumer brands have become more active stewards of sensitive customer information as traditional ways of targeting and measuring digital campaigns, such as third-party cookies, fall out of favor or encounter tougher regulatory scrutiny. Some categories, namely retail, are also building out digital ad networks that rely on shopper data, raising the stakes if that data is mismanaged. 

Failure to properly handle customer information, or worse, falling victim to a hack or breach could harm consumers and deliver a “sledgehammer” blow to perceptions of a brand, the report stated. On the flip side, a brand that is known for installing additional safeguards around data security could improve its standing.

“A strong marketing-security partnership preserves brand reputation in an environment rife with privacy concerns, proving a strong security commitment can also help build the brand,” said Donovan Neale-May, executive director of the CMO Council, in a statement. “Conversely, a weak partnership can lead to data disasters which will erode brand reputation as well as customer and employee trust.”

CMOs generally understand the value of data-security relationships, but collaboration often only occurs when there’s a fire to put out. In more nascent data-security partnerships, collaboration on campaigns sometimes does not occur at all. 

More regular data-security hygiene helps prevent crises and keeps CMOs aware of where the next threats might emerge, per the report. That said, 33% of CMOs see data-security organizations as hesitant or unwilling to work together, creating a communication barrier. 

The emergence of generative AI tools like ChatGPT was highlighted as a chief area of concern. Marketers have taken a cautious approach to AI but see it as a way of potentially boosting efficiency, creative ideation and personalization. How AI acquires and applies data can be opaque and even run into legal hot water. 

“While marketing departments want to use AI, customer behavioral data, and the Internet of Things to personalize customer interactions, they pose significant security risks,” said Bret Sanford-Chung, managing director of U.S. marketing consulting at KPMG, in a statement.

The CMO Council and KPMG’s findings were based on surveys of over 256 marketing leaders across industries and geographies, as well as qualitative executive interviews.

NEW YORK — Diageo has significantly increased its investments in artificial intelligence (AI) for precision marketing, which has helped the company navigate challenges around regulation and generating return on investment (ROI), an executive said during an Advertising Week New York session on Oct. 8, 2024. In addition to detailing the alcohol giant’s AI investments, panelists discussed how data and AI are closely related.

“We're currently building out our data capabilities, especially starting from marketing,” said Joshua Nafman, vice president, data and operations at Diageo. “How do we make better decisions? How do we drive down costs? How do we increase effectiveness?” 

Nafman was joined by Mark Zagorski, CEO at DoubleVerify. The two discussed AI implementation, the tech’s place in brand safety and the growing importance of data for integrated marketing. 

Addressing challenges

While nearly all advertisers face some sort of challenge when it comes to the digital marketing landscape, Diageo faces challenges unique to the alcohol segment, including regulatory limitations surrounding where and who it can advertise to. The company needs to reach its market without appearing next to any content aimed at those below the legal consumption age for alcohol, a problem it refers to as inventory optionality. 

“We can't market, and wouldn't want to market to anybody under the legal drinking age,” said Nafman. “So that means who we can communicate to and how we can do that…we're really challenged on that.”

Another challenge Diageo faces, like many other advertisers, is media inflation. More channels can lead to increased costs. At a time when marketing budgets face mounting pressure, figuring out how to cut costs, not raise them, is key. As marketing channels continue to splinter and diversify, that goal becomes more difficult to achieve.

A challenge specific to AI that Diageo had to address when implementing the new technology is the hype and expectation around AI, according to Nafman. Investing in technology can be extremely expensive. Data can be misused and end up not delivering the return expected of it. 

“As soon as you touch data or AI, my expectation is that my media costs will actually skyrocket…my ROI just isn't there. And we've made a lot of mistakes doing this earlier on,” said Nafman.

Nafman’s comments underscore broader concerns that have arisen around AI, in that a lot of money has been invested in the technology so far and now the pressure is on to show that it can drive results for businesses. 

Tangible returns

Despite early challenges and concerns with AI, Diageo has started to see significant returns when it comes to the technology, according to Nafman. These include increased ROIs and improved data efficiency. Specifically, the technology has allowed for hyper-specific targeting, allowing for very specific investments to be made. This is of special importance to Diageo, which has both affordable and premium products under its umbrella. 

For example, Diageo could pinpoint a ZIP code inclined toward vodka. The company would then have to choose between promoting its more affordable option, Smirnoff, or its premium product, Kettle One. In some cases, the data may support choosing to focus on Kettle One in an area, as the brand has a higher profit margin, according to Nafman. 

While much of the panel discussion focused on AI and data for marketing implementation, Diageo has also used the tech for buzzy consumer experiences. In 2024, the brand launched an activation in Edinburgh, Scotland, which allowed consumers to use AI to design their own Johnnie Walker Blue Label art and label.

Ultimately it comes down to how a company uses and visualizes its data. Once the technology has proven effective at cutting costs, marketers can take it one step further. Getting creative with data and AI can prove to have powerful results. 

“If you think about Diageo as not a spirits company, but as a data company, we're probably one of the largest holders of alcohol data in the entire world,” said Nafman. “So then it goes, well, how can I put that to work?”

Volkswagen of America partnered with Google in the fall of 2024 to bring the tech company’s Gemini and Vertex generative artificial intelligence (AI) capabilities to the vehicle owner experience, according to a press release.

The automaker is deploying Google’s generative AI capabilities in its myVW owners app to power a myVW Virtual Assistant and provide vehicle owners with smoother access to critical vehicle information and services.

The generative AI capabilities are available for Volkswagen’s My24 Atlas and Atlas Cross Sport vehicles. Using the AI-powered app, vehicle owners will be able to search their owners’ manuals by simply posing a question, such as “How do I change a tire?”

Volkswagen’s integration of generative AI into its app could help improve the overall customer experience, generating brand loyalty and an avenue to learn more about driver behavior. 

Manufacturer-created apps have become a major part of the modern auto-ownership experience. According to JD Power, nearly three-quarters of traditional, internal-combustion car owners use original equipment manufacturer apps to accomplish tasks like remote start, locking and unlocking their cars, addressing diagnostic problems and interacting with dealerships. That percentage goes up to 90% for electric-vehicle owners.

Volkwagen’s enhanced app uses Google’s AI tools, including Google Cloud’s Vertex and BigQuery, to create Gemini models for multiple data sources, including owners’ manuals, frequently asked questions, help center articles, official Volkswagen YouTube videos and step-by-step guides. The app’s interface will enable users to ask simple questions, like how to change a tire or to provide information on a dashboard indicator light. Multi-modal capabilities in the app enables users to get information on indicator lights by simply pointing their smartphone cameras at their vehicle dashboards.

“The myVW mobile app helps us advance our promise of unparalleled customer service and exceptional value for our Volkswagen vehicle owners,” said Rachael Zaluzec, senior vice president of customer experience and brand marketing for Volkswagen Group of America, in a statement. “The myVW app and Virtual Assistant effectively bring a collection of different points of information into one place, giving our owners the support they value from Volkswagen.”

As interest and investment in generative AI surges, Google has integrated its new tools through collaborations with a range of brands and agencies. Google announced a partnership with WPP earlier in 2024 that combines the tech company's generative AI tools with the agency holding company's proprietary marketing and advertising data. Previously, the Estée Lauder Companies deepened its relationship with Google Cloud to include more applications of generative AI.

Candy marketers are nearing the finish line for their category’s annual marketing blitz as people stock up on sweets for trick-or-treating. The key Halloween sales period brings both opportunities and challenges for legacy companies like Hershey that carry wide household penetration but are contending with gaps in outreach strategies that prioritize reach, executives shared during an Advertising Week New York panel on Oct. 9, 2024.

“We actually did a heat map of the United States. The entire impression load was in five U.S. cities,” said Vinny Rinaldi, head of U.S. media for CMG and salty snacks at The Hershey Company. “If you’re buying everybody with a mouth, but you’re only speaking to five cities, you’re missing everyone between LA and New York, pretty much. There’s a lot of people there that eat chocolate.”

As the maker of products like Kit Kat and Reese’s Peanut Butter Cups looks to refine its media buying and planning, artificial intelligence (AI) is helping shift focus from oversaturated markets to those that are often underserved, such as Topeka, Kansas. Hershey and fellow Advertising Week panelist HP have been working with Chalice Custom Algorithms, a four-year-old AI services provider that builds bespoke algorithms that help brands manage programmatic efforts on top digital platforms. A greater degree of precision was especially high stakes during the 2024 season amid a flood of political advertising hitting what are often quieter pockets of the country. The U.S. presidential elections were held Nov. 5, less than a week after Halloween.  

“We have to make sure we’re taking in all these different signals to make sure … we’re not driving up the value of every impression to be astronomical,” said Rinaldi. “There’s an influx of media this year, so having the algorithm built out allows us to really take all that into account so we’re not overextending during an election year.” 

High-stakes Halloween

Even in non-election years, Halloween has an outsized impact on the candy business. Failing to hit sales goals can reverberate into the months ahead and affect future holiday campaign plans. 

“Markdown day is the worst day in our business because it created waste. Not only does it create waste, but if you don’t reach certain thresholds by every single retailer, they buy less next year,” said Rinaldi, alluding to Nov. 1, when many shops apply steep discounts on unsold Halloween candy. 

Chalice, which focuses on developing custom bidding algorithms for programmatic media, sought to provide Hershey with an offering that could engage the “moveable middle” of shoppers and assist with decision-making across key platforms during the thick of Halloween mayhem, according to Chalice Co-Founder and Chief Operating Officer Ali Manning, who moderated the panel. Chalice’s tools rely on several different algorithms that work in tandem to monitor inventory availability, impression pricing and campaign pacing, among other functions. Marketing agency Lucid is a partner of Chalice.  

“We took three platforms that don’t speak together, Meta, YouTube and The Trade Desk, and built the layer above [them] to make sure that we’re allocating all of our budgets appropriately, so that we’re not over investing across each platform in those markets,” explained Rinaldi. 

Ironing out the kinks

Hershey tested Chalice for 2023’s Halloween season and outlined results with a single retailer at Advertising Week. The process could be cumbersome. Hershey had to manually share ZIP code data on a weekly basis with Chalice to see which stores it was underperforming in compared to a national average. It then took media impressions every day of those weeks to verify it had an eye on the correct markets. Despite the heavy lift, the outcomes proved promising.  

“That retailer last Halloween was the only retailer that beat sell-through,” said Rinaldi, referencing a measure that tracks the amount of supplier product a retailer is able to sell over a period of time. “And it isn’t all because of just media, but we were able to correlate the 1.5% growth on sell-through directly to media.”

“We were able to bring all these things together to actually prove the value through match-market testing,” added Rinaldi.

AI in marketing, a major topic at Advertising Week New York, continues to gain traction as an enabler of efficiency. Rather than set up thousands of line items, trafficking units and insertion orders across a trio of disparate platforms, Hershey was able to centralize more of its work and achieve a higher level of campaign “fluidity,” Rinaldi said. The next step: Applying these tactics and technologies outside of the make-or-break months of the year.  

“We’re just scratching the surface because we’re only doing this seasonally. It’s only on some of our big tentpoles. We spend a lot more every day,” said Rinaldi. “There’s a lot we can do with this and I think we’re just getting started.”

While direct-to-consumer (DTC) is no longer seen as the marketing panacea it once was — especially during the peak pandemic period — the sales channel can still be useful to CPG marketers as part of an omnichannel strategy and for collecting data that can be used to train generative AI. That is the case for Mondelēz and its flagship cookie, Oreo. 

“We approach DTC as a tool for creating closer connections to our consumers and learning from them. It enables us to test new products, marketing strategies, engagement tactics and pricing models,” said Kamila DeMaria, director and general manager of DTC at Mondelēz, via email.

To enhance its capabilities, Oreo in 2024 chose digital commerce agency Goodness as its DTC agency partner. The partnership evolved out of Goodness’ previous work to migrate Oreo from Adobe’s Magento e-commerce platform to Shopify to improve efficiency and user experiences.

“Shopify, and the tools that it unlocks, lets a brand like Oreo behave more like a digitally native brand and compete on that level,” said Ethan Martin, strategy director at Goodness. “That’s what the Oreo team specifically was after, and what Mondelēz is hoping to unlock across the whole portfolio.”

Uniting DTC and equity

Goodness’ work was two-fold: building a new e-commerce system that could be scaled across Mondelēz’s portfolio that would be efficient enough to outweigh the high cost of going direct to consumer and opening up new opportunities for the brand’s OreoID customization platform.

“[With OreoID] we really had our chance to optimize, make something that was more modern, fun, sleek and cool and that feels like a dashboard, and really try to greatly reduce customer service calls and greatly increase conversion rate for every step,” said Donald Fierros, chief operating officer at Goodness.

The work also allowed Oreo to sell retail packs and limited-edition flavors through its e-commerce shop for the first time, opened up the ability to better integrate brand campaigns and unlocked the possibility for a deeper partnership with Goodness.

“There was opportunity with Goodness’ skill sets and the match between DTC and equity sides of Oreo where their goals are different,” Fierros said.

Experimenting with data

Since coming on as an agency partner in Q2 of 2024, Goodness has worked to connect internal systems that it hopes will eventually allow for Oreo to deploy fully personalized communications that enhance engagement and conversion for DTC. Months into its relationship with Oreo, Goodness has iterated on OreoID and is working to connect ad, media and CRM data to the e-commerce site. And despite its legacy bona fides, Oreo sees itself as an innovator around data in the space.

“The advantage that Oreo has is that they have this massive pillar of retail sales that's the core of their business,” Martin said. “It gives them real room to experiment and leverage their DTC platform in a way that maybe a digital brand can’t.”

Building and experimenting with a more connected data infrastructure through DTC will be key to Oreo’s needs on the equity side of its business around how it captures zero- and first-party data at every consumer touch point. Those imperatives are likely to persist, no matter what Google’s plans for third-party cookies portend.

“Building direct relationships with consumers through compelling value exchanges is at the very core of Oreo’s marketing transformation agenda,” said Hamish Stacy, senior manager and U.S. Oreo consumer data lead, via email. “It sets the foundation for developing a sustainable competitive advantage when applied to empathy at scale across our paid and owned ecosystem.”

That data will also be key in training the company’s generative AI infrastructure, Stacy explained. Mondelēz International in September 2024 announced the launch of a new platform supported by Accenture and Publicis Groupe that seeks to better harness the emerging tech.

Understanding conversion

For marketers, collecting data is one thing. Making sure the data is valuable, understanding what has been collected and generating insights is a more complicated proposition. Goodness is helping the Mondelēz data team ensure that data flows in one direction, in the same funnel and into the same storage.

“That way, we don't have to rely on these big third-party tools to understand the data. We can understand the data on our own, and so can the teams at Mondelēz,” said Fierros. 

CRM has increasingly become a focus for Oreo, as it has throughout the marketing industry. Goodness is helping Oreo think about what first-party data it can gain by asking customers about their favorite products, limited editions and campaigns. The agency is also supporting Oreo’s efforts to determine which zero-party data it can glean from its online platforms (e.g. which content themes best drive web traffic). If done correctly, all of this data can help inform media strategy.

“Our conversion metrics won't necessarily be sales. They will depend on the challenge, like list growth or data enrichment,” Fierros said. “We've got a few North Star metrics that we're driving to, but we do have to consistently look at attribution methods to make sure that each one of those North Star metrics is being tracked in a consistent way.”

For Oreo, building and operating a stronger data infrastructure will help it support its ambitious goals around acquisition, activation and analysis that Goodness can help it achieve, Oreo’s Stacy explained.

“To reach our ambition we heavily rely on the expertise of agency ecosystem partners and their ability to work collaboratively with each other and constantly challenge, break, and rebuild ways of working,” the executive said.

The growing use of artificial intelligence (AI) in the media industry has drastically changed the skills for which industry leaders are looking. Seventy-two percent of media leaders say they have to reshape the talent and skill of their organizations to keep up with a changing media landscape, per a September 2024 report from Kantar. This is leading to changes in recruitment practices, with 66% of those surveyed saying they feel it's important to recruit outside of the media industry to meet these changing needs.

“There was an acknowledgment… telling us that they think they need to substantially reshape skills in their organization,” said Gary Brown, lead of brand communications at Kantar.

The study by Kantar Media evaluated results from 1,110 respondents across 53 markets. All respondents are involved in the media industry and have some level of data involvement.

Skills, updated

Despite the buzz, AI ranks fairly low as a desired skill among those in the media industry, with 27% of respondents saying it is a very important skill. This is in sharp contrast to communications and storytelling skills, which 79% rank as very important — more than any other skill — according to the study.

However, this doesn’t tell the whole story. While the demand for AI talent remains fairly low, demand is likely to continue to grow. Nearly 70% of respondents predict AI skills will be very important to have within the next three years. Similar tech-based skills, such as handling large data sets, and data science skills as a whole are expected to follow a similar trajectory.

“In terms of which of the skills that are going to increase in importance over the next three years? That's where you see that skills gap,” said Brown.

While AI skills are expected to be a bonus for incoming recruits, many organizations are looking internally for AI development. Over a quarter (26%) of respondents state they would like to grow skills in-house, while an additional 14% would prefer to recruit talent in-house. Only 23% of respondents would prefer to source AI talent externally, per the report. However, 32% of respondents indicated that they prefer a mix of approaches.

Industry shakeup

The media industry seems desperate for change, with 90% of respondents saying they want to recruit people who will challenge what companies are doing, while 66% feel it is important to recruit talent from outside of the media industry, according to the report.

Ultimately, findings support an industry desire for change, with a clear focus on diversifying and updating talent pools. Developing this talent goes beyond recruitment, with many respondents indicating plans for in-house training. Of the organizations planning to develop talent in-house, 63% of respondents plan on doing so through training. This is followed by eLearning, with 39% of respondents planning on implementing this method.

“There's no consensus out there as to what organizations should be doing,” said Brown. “We see this as being very much a conversation starter, right? We're trying to help diagnose the problems, but we're not coming up with all the solutions, because they're going to vary depending on who you're talking to.”