Entrepreneurs encounter significant challenges in translating their ideas into tangible or digital products. Conventional idea-to-market processes often prove inflexible, siloed, and linear, leading to missed optimization opportunities and substantial material and labor waste. To address these challenges, this paper proposes DG-ViTh, a Decision Guidance System for search, optimization and trade-off analysis of Virtual Things, which are parameterized specifications of products, services or anything of value. Virtual things are associated with analytic performance models, that express customer-facing characteristics of a "thing" in terms of virtual thing’s parameters. DG-ViTh allows virtual things’ parameters be optimally instantiated and then ordered by consumers from capability providers on demand. To develop DG-ViTh, this paper provides the design and formal mathematical semantics of a repository for general virtual things, core DG-ViTh functions, and a prototype implementation demonstrating optimization capabilities for arbitrary pluggable virtual thing specifications. DG-ViTh versatility is demonstrated using the example virtual bike, illustrating its potential for a broad range of applications.