The relentless pursuit of deeper customer understanding has long been a cornerstone of corporate strategy, and the advent of generative artificial intelligence (GenAI) has ushered in a transformative era for how enterprises approach this critical function. Businesses across sectors are increasingly integrating GenAI tools, powered by sophisticated large language models (LLMs) and techniques like retrieval-augmented generation (RAG), to unlock previously inaccessible insights from their vast reservoirs of internal and external data. This hybrid approach allows companies to fuse proprietary customer intelligence with the expansive general knowledge base of LLMs, promising unprecedented efficiency and depth in discerning market sentiment and consumer preferences. Yet, while the technological promise is immense, the journey towards truly insight-driven organizations remains fraught with persistent human and systemic challenges, mirroring obstacles that predate the current AI revolution.
The contemporary business landscape is characterized by an explosion of data, with customers interacting across a multitude of touchpoints, from direct sales engagements and customer service channels to social media, website behaviors, and extensive market research studies. This generates a colossal volume of both structured data—such as quantitative survey results, sales figures, and satisfaction ratings—and unstructured data, including interview transcripts, focus group discussions, customer emails, and social media commentary. Navigating this ocean of information to extract actionable intelligence is a daunting task. GenAI offers a compelling solution, empowering employees to access, summarize, and synthesize this content using natural language queries, thereby democratizing access to knowledge that was once fragmented across departments or buried in siloed archives. For large, complex organizations where the provenance and location of critical insights are often obscure, this capability alone represents a significant leap forward.
However, the efficacy of GenAI extends beyond merely cataloging and retrieving information. A critical distinction must be drawn between simply storing knowledge and facilitating its dynamic flow. Many companies in earlier generations of knowledge management (KM) systems, such as Lotus Notes or Microsoft SharePoint, fell short by focusing too narrowly on static repositories. While these platforms provided greater access, they often failed to instigate a revolution in insight utilization due to deep-seated cultural issues: difficulties in organizing and retrieving information, organizational silos that prevented cross-functional sharing, indifference to existing content, and a lack of seamless collaboration with external agencies. Today, despite GenAI’s advanced capabilities, these fundamental cultural and operational barriers continue to impede the full realization of its potential.
Leading consumer-oriented companies, including industry giants like Procter & Gamble, PepsiCo, and Novartis, are at the forefront of experimenting with GenAI in their customer and market insight operations. Their experiences highlight a spectrum of implementation strategies, from building bespoke in-house GenAI solutions to leveraging specialized external vendor platforms, or adopting a hybrid model. P&G, for instance, utilizes vendor-supplied software for knowledge storage and access but has developed its own GenAI system for sophisticated content analysis and categorization, enabling employees to receive "sharp, pointed answers" rather than just links to documents. This bespoke analytical layer allows for a more direct and actionable engagement with the underlying data.
Vendors in the GenAI space are diversifying their offerings to address various stages of the insight lifecycle. Some concentrate on advanced storage and access functionalities, moving beyond simple virtual filing cabinets to offer automated content curation, integration of disparate data types, on-demand analysis, and synthesized answers to complex prompts. These tools are invaluable for making pre-existing insights readily discoverable. Others specialize in the analysis of qualitative customer data, transforming historically labor-intensive processes. Still others focus on rapid consumer response testing for advertising and product concepts. The trend suggests an inevitable convergence towards comprehensive customer insights platforms that will integrate GenAI and other advanced capabilities to manage the entire spectrum of insight creation, curation, and dissemination.
A common refrain among insights leaders is that "AI is only as useful as the data it learns from." This underscores the paramount importance of data quality, curation, and robust governance. Organizations must centralize as much customer and market content as possible, diligently curating it to eliminate redundancy, obsolete information, and maintain overall quality. GenAI tools excel at categorization, summarization, and content tagging, often employing predefined taxonomies to enhance retrievability. Novartis’s "Sherlock" system for its consumer business exemplifies successful GenAI implementation for insight storage and access. Developed with an external vendor, Sherlock allows users to pose questions and receive precise answers, often pointing to specific text lines or video timestamps. Integrated "Knowledge Zones" provide expert-curated content on niche topics, while strict governance ensures document quality. By enabling research vendors to upload deliverables directly and preventing redundant research efforts, Sherlock delivered over $29 million in primary market research cost savings in a single year, simultaneously democratizing information and preventing overgeneralization of insights.
The analysis of qualitative data, traditionally a labor-intensive and often semi-manual process involving spreadsheets and manual coding, presents a particularly challenging but ripe area for GenAI. While some academic critiques suggest generic AI chatbots are unsuited for nuanced qualitative analysis, specialized GenAI tools offer a powerful alternative. Tracy Tuten, a leader in qualitative research at Illuminas (now part of Radius Insights), has pioneered "conversational qualitative data analysis." Utilizing GenAI-based software, Tuten employs natural language prompts to analyze audio and video files, generating automatic transcriptions, summarizing key themes, and comparing them across audience segments. This approach drastically reduces analysis time—a global study that once took six weeks can now be synthesized in a day—and helps uncover secondary insights that might otherwise be overlooked. This augmentation of human researchers, rather than replacement, significantly enhances both efficiency and rigor in qualitative insights generation. Similarly, PepsiCo’s “Ask Ada” platform leverages GenAI to understand customer responses to advertising and brand messaging, streamlining market research and reducing reliance on external agencies, thereby creating substantial operational efficiencies and cost savings.
Despite these transformative capabilities, GenAI is not a panacea, and it cannot supplant the strategic acumen of human marketing leaders. As Stephan Gans, former chief customer insights and analytics officer at PepsiCo, aptly noted, while "raising the bar on marketing and innovation effectiveness to fuel commercial excellence will become increasingly automated," leading the understanding of consumer demand remains a deeply strategic endeavor requiring human insight and leadership. Several critical factors, deeply rooted in organizational culture and structure, continue to pose significant challenges that GenAI alone cannot solve.
First, the lack of common approaches across geographical and business units remains a formidable barrier. A global consumer goods company, for example, might operate in over a hundred countries, each with a high degree of autonomy. Without a company-wide consensus on taxonomies for brands, product categories, or distribution strategies, knowledge becomes fragmented and incoherent. Different units invest in disparate initiatives, leading to contradictory data and outputs that are contextualized differently. In such scenarios, a GenAI tool, however sophisticated, will struggle to synthesize a unified view of global customer insights, rendering cross-learning impossible. This highlights the critical need for senior executive leadership to drive standardization and create common information formats, as exemplified by PepsiCo’s "one nation" approach under Gans, which fostered a Global Insights Council to integrate insights across regions.
Second, the absence of an insight-driven culture can render even the most advanced technologies ineffective. If an organization’s decision-makers are not ardent consumers of customer and market intelligence, market research efforts, regardless of the tools used, will fall flat. Cultivating a "passion for data" requires deliberate change management. P&G, with its century-long legacy of market research, embodies this culture, viewing consumer understanding as central to its strategy of delivering superior product experiences. This deep-seated commitment ensures that insights are not merely collected but actively sought, analyzed, and integrated into every strategic decision, from product development to marketing campaigns.
Third, the complex dynamics of agency relationships often introduce ambiguities regarding data ownership and analysis strategies. Many companies rely heavily on external advertising and marketing agencies for consumer research. This can lead to dysfunction if the client company does not retain full ownership of the data and insights generated on its behalf. Outsourcing the learning process can prevent internal employees from applying lessons learned in future campaigns and building proprietary competitive intelligence. Leaders like Gans advocate strongly for client ownership of all research results, emphasizing that while agencies provide valuable services, the strategic learning derived from the data must reside within the client organization. Collaborative software platforms can facilitate agency involvement while ensuring client data ownership and control.
Finally, the historical perception of analytics professionals as low-status "order-takers" can significantly undermine the value of customer insights. In organizations where the insights function is viewed merely as a library-like service, budget cuts and a lack of engagement are common. Users might not contribute to the knowledge base or formulate high-quality prompts, leading to underutilization of even GenAI-enabled tools. The transformation at PepsiCo under Gans serves as a powerful counter-narrative. By elevating the customer insights and analytics function to a strategic partner, deeply integrated into innovation work and supported by robust platforms like Ask Ada, the organization moved beyond an "order-taking" mentality. This strategic shift earned the function respect, increased funding, and ultimately made the company more customer-centric and effective.
In conclusion, GenAI tools are powerful augmentations, not replacements, for human expertise and well-established organizational practices. The companies that will truly thrive in the evolving landscape of customer insights are those that recognize GenAI’s capacity to enhance efficiency, democratize access, and deepen analytical capabilities, while simultaneously addressing the foundational human and organizational challenges. This necessitates proactive leadership in standardizing data, fostering an insight-driven culture, securing data ownership, and elevating the strategic role of insights professionals. The ongoing evolution of AI-enabled software will undoubtedly bring more sophisticated platforms, but their ultimate success will hinge on a holistic integration of technology, people, processes, and a pervasive culture that genuinely values and acts upon customer understanding.
