AI and the hidden climate cost of ‘dark data’
Training and fine-tuning AI models requires troves of data, but storing and processing that data is financially and environmentally costly. Here are three ways to address the problem....
by Amit M. Joshi Published 11 October 2022 in Technology • 7 min read • Audio available
Is AI ready to take over your marketing outreach to potential customers? Facebook’s owner, Meta, recently launched its BlenderBot 3 AI chat tool, yet so far the suggestion is that AI is still some way from completely taking the reins from your marketing team. Journalists quickly found the tool expressing some dubious views—not least about its parent company—telling the Wall Street Journal that Facebook has “a lot of fake news on it these days,” while asserting that Mark Zuckerberg’s business practices are “not always ethical”.
And yet the march of AI in marketing continues apace. Salesforce’s latest State of Marketing report found that 64% of marketers use AI. The demand for AI in marketing has been estimated at $12bn worldwide, and could rise to $108bn by 2028 [source: Statista]. A McKinsey analysis has identified marketing as one of the business areas that could benefit the most from AI. They estimate it could generate up to $2.6tn added value to marketing and sales. The challenge for CMOs is how they can use AI to secure their organization a slice of that value.
Some firms already use AI extensively. Common applications include automating processes, including targeting emails or online advertising, and creating ad content. For example, JPMorgan Chase has used the AI tool Persado to improve the wording used in its direct response emails and online display ads, honing in on word choices that deliver better customer responses.
Many organizations also use AI to sift customer data, seeking improved insights, a more detailed understanding of customer lifetime value (the revenue associated with a customer over the course of their entire relationship with the company), and more nuanced segmentation. As firms build a better picture of their customers, they can offer personalized recommendations and new products, much like Starbucks does with data gathered through its rewards program and mobile app. Also interesting is the use of AI and machine learning to support sales teams, for example in B2B companies, to identify next best actions: when to contact your best customers, the mode of contact, and what type of messages might be most effective.
These applications are delivering real value for businesses. However, for most companies, the current use of AI in marketing is generally ad hoc and piecemeal. I have yet to see AI and machine learning deployed in marketing in a truly well-thought-out and strategic fashion. These are game-changing technologies, but they are presently being used to improve execution on familiar marketing strategies. Using them to their full potential would mean ripping up and replacing old strategies—and that potential remains under-explored.
In practice, the bigger opportunity for CMOs lies at the point where marketing and sales meet strategy. If strategy is about where to play and how to win, the real value for marketers lies in determining where a business should compete—and, equally importantly, where it should not.
CMOs should consider how they can use AI and machine learning to discover new markets or new product segments, or to pinpoint markets they would be better served by avoiding. For example, AI could lead a company to right-size its product or service portfolio, driving simplification if machine learning revealed that a streamlined offering was optimal. Another example might be to completely change the types of customers the organization targets. This is beginning to happen, but marketers are so far only scratching the surface.
As AI becomes more deeply embedded, it will also help more businesses identify the best channels for advertising and support radically enhanced accountability for marketing spend, enabling CMOs to measure the value of brand building and campaigns to a level of accuracy that historically has been incredibly difficult: distinguishing between various contact points to understand what truly influences customers’ purchasing decisions.
AI might also help overcome the erosion of trust by fake news and false product reviews, and it has the potential to enhance diversity, allowing marketers to take account of the views of all their potential customers—not just those who the CMO understands instinctively due to a similar background or shared culture.
Improved insight, accountability, and influence could help CMOs earn a seat at the top table when it comes to key business decisions—the absence of which has long been lamented by CMOs and marketing thinkers.
For CMOs, one of the most powerful potential applications of AI could be in dynamic pricing. Can marketers more accurately discern where they can or should charge more—and when they should charge less—in a way that takes account of the lifetime value of a customer, rather than pricing only for a single transaction?
Take the aviation sector, for example. Could an airline tell if somebody browsing for a last-minute ticket is flying to do a multi-million-dollar business deal—in which case they will probably happily pay substantially more than usual—or whether they are traveling for a family emergency—in which case, perhaps the company would prefer to charge less? The dynamic pricing that currently exists uses basic data such as the time of day of browsing, or the location. But what if machine learning has identified that the passenger is likely to have a high number of journeys on a given route over the next few years: should the airline charge less to secure that additional lifetime value? An improved picture of customer behavior might also point towards enhanced service for loyal customers.
This sort of capacity does, of course, raise ethical as well as practical questions. Would the airline want to adjust prices based on an AI determination that a passenger was likely to be traveling for a family emergency? What about privacy: does the company even want to infer such knowledge? But if those conversations can be resolved, AI and machine learning could enable marketers to develop truly dynamic pricing models that are both responsive and fair to each individual customer and help generate more value over a customer lifetime.
How then can CMOs move to realize value from AI in a more strategic way? There are four key areas to consider.
Misfiring chatbots aside, it is clear that the use of AI in marketing is here to stay—but most marketers are only scratching the surface of its true potential. Learning to prioritize, scale, and orchestrate is essential if CMOs are to deliver on its true value.
Professor of AI, Analytics and Marketing Strategy at IMD
Amit Joshi is Professor of AI, Analytics, and Marketing Strategy at IMD and Program Director of the Digital Strategy, Analytics & AI program, Generative AI for Business Sprint, and the Business Analytics for Leaders course. He specializes in helping organizations use artificial intelligence and develop their big data, analytics, and AI capabilities. An award-winning professor and researcher, he has extensive experience of AI and analytics-driven transformations in industries such as banking, fintech, retail, automotive, telecoms, and pharma.
20 hours ago • by José Parra-Moyano, Christine Legner, Konrad Schulte in Audio articles
Training and fine-tuning AI models requires troves of data, but storing and processing that data is financially and environmentally costly. Here are three ways to address the problem....
18 April 2024 • by Quentin Gallea in Audio articles
GenAI will allow human talent to focus on more creative, complex tasks, leading to greater job satisfaction and well-being....
15 April 2024 • by Annabelle Gawer in Audio articles
AI-as-a-Service will impact executives and their strategic decisions. It’s crucial to assess the benefits and repercussions it will bring....
9 April 2024 in Audio articles
When the going gets tough, the behavior and decisions of board members can heavily influence the fate of companies and have far-reaching repercussions. When the going gets tough, the behavior...
Explore first person business intelligence from top minds curated for a global executive audience