PersonaBOT: Bringing Customer Personas to Life with LLMs and RAG

TL;DR

This work tackles the scalability challenge of customer persona development by generating synthetic personas from public success stories and integrating them into a Retrieval-Augmented Generation (RAG) chatbot for Volvo Construction Equipment. It systematically compares Few-Shot and Chain-of-Thought prompting for persona generation and demonstrates that Few-Shot yields more complete personas while Chain-of-Thought offers greater efficiency. Augmenting the chatbot’s knowledge base with synthetic personas and segment information yields measurable gains in accuracy ($6.42$) and perceived usefulness (81.82% rating at least somewhat useful). The study highlights practical trade-offs between completeness and efficiency in prompting methods and shows that data- and prompt-driven KB augmentation can meaningfully improve decision-support in industrial settings, guiding future work on broader data integration and advanced RAG architectures.

Abstract

The introduction of Large Language Models (LLMs) has significantly transformed Natural Language Processing (NLP) applications by enabling more advanced analysis of customer personas. At Volvo Construction Equipment (VCE), customer personas have traditionally been developed through qualitative methods, which are time-consuming and lack scalability. The main objective of this paper is to generate synthetic customer personas and integrate them into a Retrieval-Augmented Generation (RAG) chatbot to support decision-making in business processes. To this end, we first focus on developing a persona-based RAG chatbot integrated with verified personas. Next, synthetic personas are generated using Few-Shot and Chain-of-Thought (CoT) prompting techniques and evaluated based on completeness, relevance, and consistency using McNemar's test. In the final step, the chatbot's knowledge base is augmented with synthetic personas and additional segment information to assess improvements in response accuracy and practical utility. Key findings indicate that Few-Shot prompting outperformed CoT in generating more complete personas, while CoT demonstrated greater efficiency in terms of response time and token usage. After augmenting the knowledge base, the average accuracy rating of the chatbot increased from 5.88 to 6.42 on a 10-point scale, and 81.82% of participants found the updated system useful in business contexts.

Figures (16)

• Figure 1: Research Method
• Figure 2: Example Synthetic Persona - Few Shot
• Figure 3: Example Synthetic Persona - Cot
• Figure 4: Persona Generation Process
• Figure 5: Distribution of the accuracy rating