Virtual Vibes

In recent years, virtual characters have emerged as a transformative force in digital marketing. From social media influencers to online sales advisors, these digital personas are providing brands with innovative ways to engage audiences, combining the human touch with the scalable efficiency of AI. But what is actually behind their success, how exactly do brands use the artificial personalities, and what influence does their design have?

We discussed these questions with Dr. Carolin Kaiser, Head of Artificial Intelligence at NIM and lead researcher on a study exploring how virtual character design affects consumer behavior.

Virtual influencers have become a huge trend in influencer marketing. Why are they so important to a brand’s social media strategy?

CAROLIN KAISER: Virtual influencers have shown remarkable effectiveness, offering engagement rates up to 3.5 times higher than traditional human influencers. With around 300 virtual influencers on Instagram alone, and their numbers growing rapidly, they’ve become central to many brands’ social media strategies. Popular virtual influencers like Miquela and Noonoouri, for example, have millions of followers and collaborate with major brands, especially in fashion, helping brands connect with a highly engaged audience.

Why do you think virtual influencers achieve such high engagement compared to human influencers?

Virtual influencers bring a unique blend of creativity and control. Since they’re fully designed by brands or creators, they can be tailored to perfectly match a brand’s image and communicate consistently without the unpredictability that can sometimes come with human influencers. On top of that, they offer novelty—people are intrigued by interacting with digital personas, and the storytelling can feel more innovative and immersive, which keeps followers engaged.

Virtual advisors are another application of digital characters. How are they being used, and what benefits do they offer to brands?

Virtual advisors are becoming increasingly important in e-commerce. These digital humans are integrated into websites to provide personalized, interactive customer experiences, helping bridge the gap between online and in-store shopping. They add a “human element” to the digital shopping experience, increasing engagement and conversion rates and reducing cart abandonment. Brands across industries like telecommunications, finance, and retail are already using virtual advisors to boost customer online experiences.

Virtual characters offer vast creative freedom. How do brands approach the challenge of designing virtual influencers or advisors to suit their needs?

That creative freedom is both an advantage and a challenge. Virtual characters can take any form—robots, animals, cartoon characters, or even aliens—and they don’t age or face physical limitations. The real challenge for brands is deciding whether to create a humanlike or more cartoonish, nonhuman character. Both can be successful, but it’s crucial that the design aligns with the brand’s identity and the type of interaction they want to have with consumers. While humanlike characters can add realism, they risk entering the “uncanny valley,” where they might make people uncomfortable. On the other hand, cartoonish characters are often more approachable and engaging. The right choice depends on the brand’s goals and the context in which the character will be used.

This is where your research comes in. What inspired you to explore how virtual character design affects consumer behavior?

Virtual characters offer an exciting opportunity for brands to engage with customers, but it can be tricky to determine the most effective design. Many brands default to humanlike characters, assuming they’ll create stronger emotional connections. However, there are conflicting theories—some suggest humanlike characters enhance consumer engagement, while others argue they may cause discomfort. We wanted to test these theories in an online experiment by comparing humanlike and cartoonish designs to see how they influence consumer behavior in realworld marketing scenarios.

Could you walk us through the experiment you conducted?

We designed an online experiment involving 2,000 American participants, who were divided into four groups. Each group interacted with either a virtual influencer or a virtual advisor, with one experiencing a humanlike character and the other a cartoonish one. These characters were integrated into a simulated shopping journey. It began on a social media platform similar to Instagram, where participants received product recommendations from the virtual characters. Then, they visited a custom-built online store where they could potentially purchase the recommended products. Finally, participants completed a survey to share their feedback on both the character and their overall shopping experience.

What were the key findings from the experiment?

One striking result was that both types of virtual characters—cartoonish and humanlike—effectively promoted products. Across all groups, participants purchased significantly more items than we would expect by chance. However, when it came to virtual influencers, the cartoonish design clearly outperformed the humanlike version. Participants found the cartoonish influencer to be more emotionally engaging, lifelike, and less unsettling, leading to stronger parasocial relationships, higher satisfaction, and ultimately more purchases. Interestingly, in the case of virtual advisors, the design had little impact on consumer attitudes. Purchase behavior was similar for both the cartoonish and humanlike advisors, suggesting that character design plays a more critical role in influencer contexts, where storytelling and emotional connection are key.

Why do you think the cartoonish influencer resonated better with consumers compared to the humanlike one?

It’s likely due to the creative flexibility and expressiveness that cartoonish characters offer. They can vividly convey emotions and personality traits without crossing into the "uncanny valley," where humanlike characters can sometimes evoke discomfort due to their almost-but-not-quite-human appearance. The exaggerated features of a cartoonish character allow for clearer, more engaging narratives, making the influencer feel more relatable and likable to consumers.

And how do you interpret the minimal impact of character design in the virtual advisor context?

Virtual advisors serve a more functional role, guiding consumers through a store and helping with purchase decisions. In this context, the character’s design seems to matter less because the primary focus is on enhancing the shopping experience rather than building a personal connection. Since advisors provide practical information, the visual appeal or emotional impact of their appearance isn’t as crucial as it is for influencers, who rely heavily on storytelling and engagement to influence consumers.

Based on your findings, what advice would you give to brands considering virtual characters in their marketing strategies?

I would advise brands to think carefully about the context in which they plan to use virtual characters. For influencers, for whom emotional connection and engagement are vital, a cartoonish design might be more effective, as it allows for richer storytelling and relatability. On the other hand, for functional roles like virtual advisors, the focus should be on the character’s usefulness and how it enhances the shopping experience. In this case, the design is less important. A well-thought-out design strategy that aligns with the role and goals of the character can significantly enhance its effectiveness in marketing.

What broader implications does your research have for consumers and society?

One major implication is the increasing need for digital literacy among consumers. As virtual characters become more prevalent, it’s essential for people to critically assess the influence these entities have on their decisions. On a societal level, we also need to consider the ethical aspects—particularly regarding the impact of hyper-idealized virtual influencers on people's perceptions of beauty and self-worth. There are genuine concerns about the effects these perfect, digitally crafted characters might have on social norms and personal identity.

Lastly, what are the next steps for this research?

There’s still a lot to explore. Future research could examine how different demographic groups respond to virtual characters or how long-term exposure influences consumer behavior. We’re also interested in delving deeper into the ethical implications, especially as virtual characters become more integrated into our everyday lives. As the use of digital humans expands, understanding these dynamics will be increasingly important.