The modern enterprise operates amidst a deluge of data, a ceaseless flow of customer interactions, market trends, and competitive intelligence. For decades, companies have grappled with the formidable challenge of extracting actionable insights from this vast ocean of information. While traditional knowledge management systems often struggled to keep pace with data volume and complexity, the advent of generative artificial intelligence (GenAI) presents a transformative opportunity. Leveraging the sophisticated language and reasoning capabilities of large language models (LLMs), GenAI tools are empowering organizations to access, analyze, and synthesize proprietary customer insights with unprecedented speed and depth, yet success hinges critically on overcoming deeply entrenched organizational and cultural obstacles that predate the current AI wave.
Companies are increasingly adopting hybrid knowledge approaches, often employing techniques like retrieval-augmented generation (RAG), to integrate their internal customer intelligence with the broader knowledge base on which LLMs are trained. This synergy allows employees to query vast internal repositories using natural language, receiving summarized, contextualized answers rather than just links to documents. This capability is particularly invaluable in sprawling global organizations where insights may be siloed across departments, geographies, or product lines, making it nearly impossible for individuals to locate or even know of their existence. The sources of customer insights are diverse, ranging from formal market research reports and sales interactions to customer service tickets, social media sentiment, website analytics, and qualitative data from focus groups or interviews. The sheer volume and varied formats of this structured and unstructured information make automated tools for summarization, categorization, storage, and retrieval not merely beneficial, but essential.
However, a singular focus on the technological aspects of storing and accessing knowledge risks repeating past errors in knowledge management initiatives. The true power of GenAI extends beyond creating sophisticated virtual filing cabinets; it lies in its ability to enhance knowledge flows—the entire lifecycle encompassing how customer and market insights are created, analyzed, curated, disseminated, and ultimately applied. While earlier technologies like Lotus Notes and Microsoft SharePoint offered improved access, they often failed to instigate a revolution in insight utilization due to persistent cultural challenges. These included difficulties in organizing and retrieving knowledge, indifference to content, organizational silos, redundant efforts, and inadequate collaboration with external agencies. Today, these very same human and organizational factors remain the critical determinants of GenAI’s effectiveness in the insights domain.
Leading consumer-oriented companies, including Procter & Gamble, PepsiCo, and Novartis, are at the forefront of integrating GenAI into their customer insight strategies. Their experiences reveal a spectrum of approaches, from leveraging vendor-supplied software for core functions to developing bespoke GenAI systems for specialized analysis. Procter & Gamble, for instance, employs vendor solutions for knowledge storage but has engineered its own GenAI system for the nuanced analysis and categorization of content. This allows them to extract "sharp, pointed answers" from GenAI, a significant leap beyond simple document retrieval. The vendor landscape itself is rapidly evolving, with some providers focusing on insight storage and access through automated curation, integration of diverse content, and on-demand analysis, while others specialize in the complex realm of qualitative data analysis or rapid consumer response testing. The trend suggests a future where comprehensive customer insights platforms will emerge, powered by GenAI and other advanced capabilities to manage the entire insights process end-to-end.
A fundamental truth consistently articulated by insights leaders is that "AI is only as useful as the data it learns from." This underscores the critical importance of data quality, curation, and centralization. Organizations adopting GenAI for insights often consolidate as much customer and market content as possible into a unified system. Rigorous curation is then necessary to eliminate redundancies, discard obsolete information, and maintain overall data integrity. GenAI tools excel at categorization, summarization, and content tagging—tasks that, when performed manually, are both time-consuming and prone to inconsistency. Automated tagging, guided by predefined taxonomies, significantly enhances the retrievability and utility of insights.
Novartis provides a compelling illustration of GenAI’s impact on insight storage and access. Through collaboration with an external vendor, the company developed "Sherlock" for its consumer business. This system allows users to pose questions and receive precise answers, often pointing to specific text lines or video timestamps. Sherlock also incorporates expert-curated "Knowledge Zones" on specialized topics, such as packaging. Crucially, the system imposes strict governance guidelines for content contributors, ensuring high quality and consistent formatting. Research vendors can directly upload deliverables into Sherlock, streamlining the flow of external insights. This strategic implementation has yielded tangible benefits: Novartis saved over $29 million in primary market research costs in just one year. Beyond cost savings, Sherlock democratizes information, enabling employees to swiftly access relevant knowledge and identify subject matter experts, while features like "WatchOut" prevent overgeneralization by flagging geographically specific data, for example.
Qualitative data analysis, traditionally a messy and labor-intensive undertaking, represents another area profoundly impacted by GenAI. Historically, market researchers often relied on semi-manual processes involving spreadsheets and meticulous coding. While some academic critiques suggest generic AI chatbots are ill-suited for qualitative analysis, specialized GenAI tools offer robust capabilities. Tracy Tuten, a leader in qualitative research at Illuminas (now Radius Insights), adopted generative AI software to mine customer insights, coining the term "conversational qualitative data analysis." This approach enables her to upload audio and video files for automatic transcription, generate summaries, identify emergent themes, and compare findings across audience segments using natural language prompts. What once took six weeks for a large-scale global qualitative study can now be synthesized in a single day, dramatically enhancing efficiency and surfacing secondary insights that might otherwise be overlooked. This advancement significantly improves the rigor and consistency of qualitative analysis, although it serves to augment, rather than replace, the critical interpretive role of human researchers.
PepsiCo’s "Ask Ada" platform exemplifies a comprehensive application of GenAI in marketing research, particularly in assessing consumer responses to advertising campaigns and brand messages. This multi-faceted system integrates various components, including rapid concept testing, real-time sentiment analysis across digital channels, predictive modeling for market acceptance, and a centralized repository for historical campaign performance data. Ask Ada has not only accelerated PepsiCo’s insights generation but also significantly reduced its reliance on external agencies and consultants, fostering a more agile, in-house insights capability and strengthening institutional knowledge.
Despite these impressive technological advancements, GenAI is not a panacea. It cannot replace strategic human leadership, particularly in shaping marketing strategy and driving innovation. As Stephan Gans, PepsiCo’s senior vice president and chief customer insights and analytics officer, observed, while raising the bar on marketing effectiveness may become increasingly automated, "leading the understanding of consumer demand is much more strategic and still requires humans." Our research indicates that several fundamental issues continue to inhibit the full potential of AI technology in transforming customer and market insights. These are not new problems but enduring human and organizational challenges that must be addressed concurrently with technological deployment.
One significant hurdle is the lack of common approaches across geographical and business units. In a global consumer goods company we examined, a GenAI-based knowledge tool struggled to achieve widespread adoption due to the profound autonomy of country-based units. Inconsistent naming conventions for brands, categories, and distribution strategies led to "pockets of knowledge" that were "very incoherent," characterized by "contradicting numbers and outputs—all contextualized differently." Without a senior executive mandate to standardize information formats and foster greater commonality, the GenAI tool remained underutilized, preventing cross-learning among marketers and product developers. This contrasts sharply with PepsiCo’s journey, where Gans, with strong CMO support, created a "one nation" approach to market research through the Global Insights Council, ensuring tight integration of insights into innovation work across regions.
Secondly, the absence of a culture where customer and market insights are integral to strategy is a critical impediment. Even the most advanced technologies will falter if decision-makers are not ardent consumers of data-driven knowledge. This necessitates significant change management and cultural transformation. P&G, with its century-long legacy of market research, embodies this insights-driven culture. Kirti Singh, P&G’s chief analytics, insights, and media officer, emphasized, "At the heart of everything we do is the consumer. Our strategy is to provide a superior product experience… We employ experimental science, human and behavioral science, data science, and technology platform knowledge to understand our consumers." Such deeply embedded consumer-centricity creates fertile ground for GenAI to thrive.
Thirdly, agency relationships often introduce complexities around data ownership and knowledge transfer. Many companies outsource consumer research to external advertising and marketing agencies, which can lead to ambiguity regarding analysis strategies, interpretation, and ongoing ownership of data and results. If agencies retain ownership or control over most customer and market insights, the client company’s employees become dependent on external help to meet customer needs. PepsiCo’s Gans strongly advocates that client companies must own all research results and insights generated by agencies on their behalf. He argues against owning the data but outsourcing the learning, as internal teams need to apply lessons from market research to future campaigns. Effective GenAI platforms can facilitate collaboration by allowing both clients and agencies to view, edit, and query research, provided clear data governance and ownership agreements are in place.
Finally, the perception of analytics professionals as low-status "order takers" can severely undermine insights initiatives. In one consumer products company, the insights function historically operated like a library, with internal customers needing to consult a researcher. While a vendor-supplied GenAI tool democratized access, users still treated it passively, contributing little and often providing low-quality prompts. Despite the new capabilities, budgets and headcount for insights were cut, and international units resisted adopting the tool. PepsiCo faced a similar "order-taking" mentality, where researchers lacked respect and influence. Gans and the CMO recognized that investing hundreds of millions annually in insights made little sense without becoming truly customer-centric. Their transformation, including the implementation of AI and insights software like Ask Ada, elevated the customer insights and analytics organization to a respected, well-funded strategic partner, shifting its role from reactive reporting to proactive collaboration and strategic guidance.
In conclusion, GenAI tools are powerful augmentations, not replacements, for existing organizational strengths. As P&G’s Singh notes, their approach involves "bringing together our tradition of being focused on understanding the customer with the latest AI solutions—not replacing, for example, customer home visits, but augmenting them with AI." The rapid evolution of AI-enabled software for managing customer insights offers diverse features, but leaders must recognize its inherent limits. In highly decentralized global organizations with inconsistent data taxonomies and fragmented approaches, even advanced AI struggles to synthesize a unified view of customer and market insights. Human intervention is indispensable for integrating, standardizing, and contextualizing data to ensure effective analysis and action. Ultimately, no software, however sophisticated, can cultivate an interest in gathering and acting on customer insights if that passion is absent within the workforce. The successful integration of GenAI into customer intelligence strategies demands a holistic approach, blending cutting-edge technology investment with profound organizational redesign, cultural transformation, and visionary leadership to thrive in an increasingly data-rich and customer-centric global economy.
