Skip to Content Skip to Footer
Reality Check: AI Will Probably Not Solve Your Biggest Insights Challenges

Reality Check: AI Will Probably Not Solve Your Biggest Insights Challenges

Joris Zwegers and Paul Ricciuti

white and black robots that look like humans standing in rows

Artificial intelligence (AI) has marketers buzzing—but across industries, marketers and insights professionals are struggling to successfully leverage AI.

Knowing that effectively distilled and deployed insights act as the seeds of growth,1 the AMA and Kantar surveyed 606 marketers to understand where and how AI is used in insights development today, what roles it may displace, and, in turn, what skill sets organizations should prioritize to maximize the benefit created by insights professionals and AI tools.


The research reveals that while there is broad excitement for AI and a sense that there are specific actions AI can perform effectively, marketers feel AI is actually less suited to solve the top challenges facing insights teams today: crafting insights-backed stories and activating new ideas with relevant business leaders.

Insights Development: Key Challenges

The insights development process comes in many shapes and forms but broadly follows a similar pattern after a problem is clearly defined:

  1. Data Assessment: Insights teams build a sense of understanding of all the data they have collected.
  2. Insight Distillation: The data is analyzed and transformed into actionable insights with the aim of answering one or more key business questions.
  3. Activation: Insights are firmly linked to business strategies and presented in compelling stories that encourage stakeholder adoption and activation.

The top three challenges insights professionals face (when sufficient, quality data is available) are not found in the data assessment phase. They’re found in the transition from insight distillation into the activation phase: connecting the dots across multiple data sources to reveal the bigger story, crafting that story, and gaining stakeholder buy-in on the result (see Exhibit 1).

Bar graph showing top challenges when distilling and actioning insights from multiple sources, ranging from 51 percent for connecting dots across sources to 20 percent for being overwhelmed with data. The three main challenges noted in the text are highlighted.
Exhibit 1

The Current Use of AI for Insights

Today, 23% of marketers state they have AI capabilities built into their insights processes and are actively using them. This usage is commonly found in the initial data assessment phase of insights development (see Exhibit 2).

Bar graph showing how AI is being used in organizations, ranging from 44 percent for generating new ideas to 9 percent for gaining stakeholder buy-in. Four prominent uses are highlighted: navigating large amounts of data, identifying points of interest in data, organizing data to be actionable, and describing data and information.
Exhibit 2

This corresponds with how marketers are currently interested in leveraging AI: while there is broad interest in using AI across almost all use cases, marketers are most interested in using AI to tackle their lower-order challenges around assessing and managing data (see Exhibit 3).

Bar graph showing levels of interest in development of AI for marketing use cases in organizations, ranging from 76 percent for navigating large data amounts to 50 percent for understanding consumers’ human side. The same four uses highlighted in Exhibit 2 are highlighted again, now at 68 percent to 76 percent.
Exhibit 3

AI Usage in the Future

While today’s AI usage and interest intersect in the early stages of the insights process, perceptions of AI capabilities indicate that there is appetite for AI to be deployed beyond data assessment, into areas where it may outperform people (see Exhibit 4).

Bar graph showing which is better at various use cases: people or AI or equal. AI is shown as best at navigating large amounts of data (78 percent); people are best at understanding the human side of consumers (84 percent).
Exhibit 4

In the future, AI may aid in the transition from insight distillation to activation. There is interest in deploying AI against a top challenge: connecting the dots across different data sources for the bigger story. While only 15% of respondents state they are currently using AI in this capacity, 72% stated they are interested in doing so, and 49% of respondents think AI is already better than people at performing this task, compared with 29% who think people are superior.

What Does This Mean for How Organizations Compete?

Insights 2030 and Insights 2020 research showed that the competitive advantage separating leading and lagging insights teams lies in the ability to transform data into actionable strategy. Leading organizations have processes and tools in place, now including AI, that free them to focus on strategic priorities and challenges rather than exerting effort on less valuable tasks.

Developing and leveraging an insight to change opinions and business strategies—activating it—is a challenge and strategic priority where marketers feel AI is ill-equipped to help:

  • 76% of respondents believe it is somewhat or extremely difficult to gain stakeholder adoption of new data or insights, and 75% feel this is a problem better solved by people over AI.
  • Similarly, even though there is some usage of AI in these areas, respondents think that telling compelling stories based on insights and developing strategies based on insights are also better accomplished by people (73% and 63% respectively).

Marketers are adopting AI to better assess their data and are hopeful that it can help them connect dots across multiple data sources. But activating insights, which includes some of the biggest, mission-critical challenges insights professionals face, remains a human skill set.

Organizations can’t remain on the sidelines as their competitors begin leveraging AI. The adoption of AI to handle lower-order challenges may further the divide between organizations. Leading companies that have already deployed AI to help with data assessment can focus their people on developing activation skill sets that drive business success. Laggers, on the other hand, must spend time on data assessment tasks, which may be important but don’t directly drive business outcomes. Regardless of your current state of business, if you’re afraid of being left behind, consider how AI can be leveraged in your own data assessment phase—the common place to start.


1. Kantar, “Insights 2030: The Imperative of Imagination,”

Joris Zwegers is a Partner at Kantar Brand Consulting

Paul Ricciuti is an Associate Director at Kantar Brand Consulting