The Human Factor in Virtual Influence

Digital marketing has been transformed by the advent of virtual characters thanks to breakthroughs in computer graphics and artificial intelligence (AI). This evolution has introduced innovative ways to interact with customers, featuring virtual influencers captivating audiences on social media or virtual advisors offering assistance in online stores.

In contrast to human characters, working with virtual characters poses no physical limitations to the character and its portrayed life. However, the advantage of unlimited degrees of creative freedom comes with the burden of choosing a design that perfectly fits a brand’s needs. One major design decision is choosing between a human-like or cartoonish appearance for a virtual character. Faced with an overload of choice, brands tend to choose human-like virtual characters as they are believed to allow followers to create a stronger bond with them and, therefore, have the power to influence their purchasing decisions. However, human-like virtual characters and assistants can also be perceived as creepy, leading to aversion, non-purchase, and negative word of mouth. There were conflicting scientific theories about the impact of human-likeness on consumer behavior, with one suggesting positive perceptions and increased sales and the other suggesting consumer eeriness and non-purchase. This research aims to shed light on the impact of the human-likeness of virtual characters on consumer reactions and purchase behavior.

For this purpose, we conducted an online experiment that simulated a shopping experience and featured a virtual character. Recognizing the widespread use of virtual characters in the United States, we invited 2,000 American participants and split them into four distinct groups that included a human-like or cartoonish virtual influencer or advisor. During the experiment, the participants first interacted with a virtual influencer or advisor, receiving product recommendations. They then visited an online store to potentially purchase these items before providing feedback through a survey. To facilitate this experiment, we collaborated with a design agency to create two similar virtual characters—one with a human-like appearance and the other with a cartoonish look. These characters were then integrated into a prototypical social media platform, mirroring Instagram, and a custom-built online store.

The research reveals that both cartoonish and realistic virtual influencers and advisors effectively promoted the products. Across all settings, the participants purchased a significantly higher number of products recommended by the virtual characters than would be expected randomly. Depending on the context, however, the design of the character led to different consumer attitudes and behaviors. In the realm of virtual influencers, character design played a crucial role. The cartoonish virtual influencer outperformed its human-like counterpart significantly. It was viewed as more lifelike, exhibiting emotions, intentions, and personalities while evoking less discomfort. The participants found the storytelling of the cartoonish influencer more engaging, developed a stronger parasocial relationship with it, expressed greater satisfaction, and were more inclined to endorse it. As a result, the participants were more likely to purchase products endorsed by the cartoonish influencer in the online store and engage more actively with it on social media.

In contrast, for virtual advisors, the design’s impact was minimal. Both the cartoonish and human-like characters achieved similar consumer attitudes and purchase behavior. This variance suggests that the influencer context, which emphasizes personal connection and narrative, is more sensitive to character design than the advisor context, which focuses more on products and shopping experiences. Our study offers valuable insights for marketing practitioners aiming to leverage virtual characters in their digital strategies. By judiciously considering design elements and the specific application context, brands can maximize the impact of their virtual personas. For consumers, enhancing digital literacy is crucial to critically evaluating the influence of these virtual entities in their lives. From a societal perspective, this research also underscores the need for ethical considerations and discussions about the influence of virtual characters to safeguard consumer interests and address potential impacts on identity and social norms.

• Overall, both the cartoon and realistic characters significantly exceeded random conversion rates for recommended products in the contexts of both virtual influencers and advisors.
• The design of the character plays a significant role for virtual influencers, with the cartoonish influencer generating more satisfaction, higher rates of recommended product purchases, and greater social media engagement.
• In contrast, for virtual advisors in online stores, the design’s impact is minimal, as both cartoonish and realistic characters foster similar consumer attitudes and purchasing behaviors.