
A technology company is developing an AI-powered platform designed to help Thai SMEs simplify accounting, tax compliance, payroll, and other back-office operations. Prior to launch, the client needed to understand how SME owners and finance personnel currently manage these processes, identify their key operational pain points, and evaluate reactions to a working prototype. The research also had to answer a harder question: whether Thai business owners would trust AI-driven financial automation at all — and under what conditions — alongside feature relevance, adoption intent, and pricing expectations.
Iconic Research recruited SME decision makers and ran ten 60-minute in-depth interviews with business owners, finance personnel, and accounting staff, drawn from businesses of different sizes across e-commerce, F&B, retail, beauty and salon, and service industries.
Each interview included structured feature testing: respondents performed seven interactive tasks simulating real workflows — company registration, revenue reconciliation, tax compliance and payment, expense management with withholding-tax automation, invoice creation via chat, payroll processing, and auditor integration. This meant reactions were grounded in hands-on use of the prototype rather than descriptions of it.
The analysis framework covered workflow assessment, pain-point identification, prototype evaluation, AI trust testing, pricing exploration, and adoption-driver analysis, with recruitment and fieldwork managed end to end by Iconic Research.
The study delivered a clear read on where the platform creates value and where adoption will meet resistance. Respondents showed strong interest in automating tax compliance, expense management, reconciliation, and payroll — the processes that consume the most manual effort and administrative cost. But the research surfaced the condition that will shape the product’s design: trust. Most users want visibility and verification before allowing AI to execute financial actions independently. For the client, that finding converts a launch assumption into a design requirement — automation with human oversight, not automation instead of it — before the product reaches the market.