Who is this research for? Sales leaders, marketing managers, and organizations using online chat channels to generate and qualify B2C sales leads.
Executive Summary
This research from Valerie Good at the Sam M. Walton College of Business, University of Arkansas (Department of Marketing), investigates how language used in online chat conversations can help organizations better identify high-quality business-to-consumer (B2C) sales leads. Drawing on three studies—including sales manager interviews, predictive text analytics of a large U.S. automotive dealership’s chat transcripts, and a replication using data from a national furniture retailer—the research examines which conversational cues meaningfully signal downstream purchase likelihood and profitability.
The findings indicate that customer language patterns, summarized in the MINITS framework—Mode of contact, Immediacy, Need, Interest, Time spent chatting, and Specificity—can serve as practical indicators of buyer intent. While not all intuitive signals used by salespeople proved predictive, several textual cues—including stated needs, mentions of specific follow-up times, and the overall duration of the chat—were associated with higher purchase probability. The research suggests that integrating conversational signals into lead-scoring processes may help B2C firms more effectively prioritize follow-up efforts when dealing with high lead volumes and limited sales resources.
Action Items for Industry
- Use chat data to enhance lead scoring: Incorporate conversational cues—not just contact info or basic engagement metrics—into existing lead-scoring systems to better predict which prospects are likely to convert.
- Train teams to recognize need-related signals: Equip sales staff to identify clear expressions of need or problem urgency, which the research shows are meaningful predictors of downstream purchases.
- Prioritize leads with time-related specificity: Give additional weight to prospects who mention specific dates or times to follow up, as this clarity significantly strengthens indications of purchase intent.
- Monitor chat duration as a signal of intent: Use the length of the conversation as an effort-based cue—longer chats were associated with higher likelihood of purchase, reflecting greater prospect engagement.
- Adopt text analytics to scale qualification: Use automated tools that detect linguistic patterns aligned with MINITS to help sales teams triage high volumes of digital leads more consistently and efficiently.
Quote from the Researcher
“By analyzing the conversational signals hidden in online chats, our findings show how AI can help salespeople quickly pinpoint which prospects are true buyers versus casual browsers—dramatically improving the efficiency of lead qualification in a digital-first marketplace. As traditional in-person cues disappear, textual signals become powerful indicators of purchase intent and potential profitability, giving sales managers actionable, real-time guidance on which leads deserve immediate follow-up.”
- Valerie Good
Co-Authors & Affiliations
- Abhi Bhattacharya — The University of Alabama, Culverhouse College of Business
- Bryan W. Hochstein — The University of Alabama, Culverhouse College of Business
- Clay M. Voorhees — The University of Alabama, Culverhouse College of Business
Link to the Original Research
Published in International Journal of Research in Marketing, available here.
📩 Interested in learning more?
If you’d like additional information about this research or to connect directly with
the researchers, please email us at research@walton.uark.edu.
