Many customer-focused companies increasingly rely on generative AI alongside large language models (LLMs) to better understand their customers and markets by analyzing internal content. These hybrid approaches, often using retrieval-augmented generation (RAG), combine company-specific insights with general LLM knowledge bases to improve knowledge management. Benefits include easier access and natural language summarization of insights, which is especially valuable in large organizations where employees struggle to locate and use customer data.
Customer insights come from diverse sources like market research, sales interactions, social media, and purchase behavior. While AI tools help summarize and categorize this vast information, focusing solely on storing insights is insufficient. Organizations must also enhance how insights are created, analyzed, and shared, overcoming challenges that plagued earlier knowledge management attempts such as organizational silos and cultural resistance.
Programs like Procter & Gamble’s GenAI system and Novartis’s Sherlock platform exemplify how AI enhances insight accessibility and reduces redundant research costs. Additionally, generative AI enables qualitative data analysis through tools that transcribe, summarize, and surface themes from interviews and focus groups, speeding up traditionally lengthy processes without replacing the expertise of human researchers.
However, AI currently cannot replace strategic human decision-making and faces limitations due to inconsistent data standards across global units, insufficient integration into corporate culture, complexity from agency relationships, and perceptions of analytics roles. Companies like PepsiCo have addressed these issues by fostering unified approaches, empowering insights functions, and emphasizing human-AI collaboration.
Ultimately, while generative AI tools can significantly augment customer insights management, they must be part of a broader cultural and strategic framework that values data-driven decision-making and effective knowledge sharing.
