ENHANCING SME CAPABILITIES WITH AI AGENTS USING APPRECIATIVE INQUIRY TO DRIVE SALES GROWTH AND REDUCE WORKLOAD IN ORDER MANAGEMENT: A CASE STUDY OF XYZ CO., LTD. (PSEUDONYM)

Abstract

This study aimed to synthesize best practices for implementing AI Agents in SME sales and order management, design an Appreciative Inquiry-based implementation framework, and pilot it through a case study with continuous reflection. Employing an Action Research design guided by the Appreciative Inquiry 4D Cycle (Discover, Dream, Design, Destiny), the case organization was XYZ Co., Ltd. (pseudonym), a micro-sized skincare and cosmeceutical distributor operating primarily through Facebook with 1-5 personnel, where the researcher served as Practitioner-Researcher alongside five key informants representing the entire workforce. Data were collected via Appreciative Interviews, participant observation, and Meta Business Insights, with a Pre-test Baseline during November-December 2025 and the 4D Cycle conducted January-March 2026) Six best practices emerged: initiating from team stories rather than technology, clearly delineating AI and human responsibilities, designing for round-the-clock coverage, employing iterative pilot testing before go-live, evaluating through multiple indicators simultaneously, and cultivating a continuous learning culture. The study also produced the AI-Appreciative Development Cycle (AADC) model. Qualitatively, team members evolved from passive users into active system evaluators capable of identifying gaps and providing constructive feedback. Supplementary quantitative data from Meta Business Insights showed sustained performance efficiency despite significantly increased workload, with Response Rate held at 100%, average response time decreasing from 2 minutes 22 seconds to 1 minute 30 seconds, Conversations increasing by 258%, Orders by 290%, and Conversion Rate improving from 10.74% to 11.70%, though these figures are not attributed causally to the AI system given concurrent advertising campaigns as uncontrolled confounding factors.

Keywords: AI Agents, Appreciative Inquiry, Small and Medium Enterprises, Sales Process, Order Management, Action Research, Information System