AI and Marketing: Cutting Through the Fog of Hype

The arrival of the “AI Age” has prompted wild speculation, expectation, consternation and more, with a fire hose of opportunities deluging CMOs. Some seem promising. But which should you choose?

Some leaders have chosen to hang back and let the competition make the big (some right, some wrong) investments. While this is a defensible approach, I don’t recommend it. Why? First, your competition is also trying to figure out what to do. AI will inevitably be a significant part of the CMO’s job. So, CMOs need to understand it—not just read about it—to remain competitive and avoid putting their business at a disadvantage.

Second, if you don’t get smart on AI now, you risk losing influence over technology decisions that are already underway. As we saw with other technologies, such as digital experience platforms (DXPs), if marketing doesn’t have a strong voice in asserting its needs and priorities, then other groups fill the void. Those viewpoints and priorities are valid and important, but they are not the same as yours. Marketing needs to be in the game, with a seat at the table, or else risk seeing its needs deprioritized.

My advice is to start planting seeds and see what grows. Identify a modest portfolio of cases that could benefit from generative AI, and run proof-of-concept projects to see what works now and what looks promising down the road.

Three areas have clear value propositions for marketing. The technologies are comparatively mature and can be implemented as beacon initiatives in about three months for a not-exorbitant investment.

ABM campaigns deliver sequenced messages to specific accounts or prospects. AI can analyze vast datasets to segment audiences based on behaviors, preferences and demographics, potentially helping marketers identify high-value target accounts more efficiently and with greater precision.

Additionally, AI can generate and implement multiple creative concepts for copy tests, while continually evaluating campaign performance and implementing adjustments to find precisely the right offers and language to increase engagement and improve ROI. AI can then turn its insights into reports on campaign performance, customer behavior and market trends, leading to better-informed decision-making across other marketing channels.

AI-powered chatbots have demonstrated an uncanny ability to parse customer inputs and communicate convincingly. They can leverage data from sources such as CRM, social media and web analytics, to create a holistic view of the customer. If given access, they can also analyze customer data and provide highly personalized and relevant responses, increasing customer satisfaction and loyalty. Their ability to learn from each interaction facilitates continual improvement in service delivery.

AI can help marketers automate and optimize many tasks and workflows. It can provide data-driven insights such as the best time, channel and format to reach various audience segments. AI can also increase the efficiency of content distribution by using programmatic placement and optimization, which can improve the reach and relevance of their ads. Finally, AI can measure and optimize marketing content performance by analyzing impressions, clicks and conversions and adjusting accordingly.

Unsurprisingly, taking advantage of AI comes with challenges. It's important to ask yourself the following questions.

Data is essential, as large language models (LLMs) analyze massive datasets, find and map connections, and then use those connections to generate responses. Thus, having relevant data to “train” the AI is a prerequisite for any AI marketing effort (i.e., targeting, customer engagement and personalized interactions).

Having data is just the start. The data must be clean—free of errors, duplication and inconsistencies that would confuse the AI and corrupt its responses. The data must be normalized and accessible to the machine-learning engine. There must be sufficient data to support the model; this is a particular challenge in many specialized B2B categories, as there is simply not the same quantity of relevant data available as in, say, popular consumer categories.

You didn’t think we could talk about AI and marketing without discussing legal and compliance, did you? No surprise, there are multiple potential legal hurdles you may encounter on the way to implementing AI. There are privacy questions (is private data being used improperly to train an LLM?). There are brand questions (is a conversational AI “on brand” in the way it engages customers?). And there are liability questions (what if an AI provides wrong answers or even right answers that lead to undesirable or unfortunate customer behavior or actions?). Given the “black box” nature of AI, all business stakeholders are rightly concerned about providing safeguards.

AI is a paradigm shift we are only beginning to grasp. There are no right answers, but there are right actions. Get smart about the technology. Get control of your customer data. And maybe most importantly, take a portfolio approach and get in the market with test programs and trials that will help you learn which are the best opportunities for you—and ensure marketing has a voice in enterprise-wide decisions.