This thesis investigates the use of autonomous sales agents in virtual stores. In the real world, specialized salespeople are often employed for customer consultation in shops. In addition to recommending products, they can also suggest cross-selling and up-selling products to customers. The motivation behind this is to encourage customers to make purchases, foster long-term loyalty to the company, and thereby increase the company's revenue or profit. Can these concepts also be applied to virtual environments? This is where the thesis comes in, as online shops and salespeople can also be implemented in virtual worlds. Virtual stores can save (personnel) costs, as products can be displayed in a comparatively cost-effective manner for the customer. The main question of this work addresses whether virtual agents can provide meaningful advice to customers. To explore this, a sales agent and a store were implemented in a virtual world simulated by Second Life and evaluated in a laboratory study.