Publications
Manewitsch, V., Kaiser, C., Biedermann, L., & Dieckmann, A. (2025). From Cuteness to Conversion. Rethinking AI in Brand Communication. NIM Insights Research Magazin Vol. 8 - AI.Meets.Consumer.
2025
Dr. Vladimir Manewitsch,
Linus Biedermann,
Prof. Dr. Anja Dieckmann
From Cuteness to Conversion
New research reveals that anime-style virtual influencers outperform realistic ones, especially in campaigns promoting sustainability and social values. Discover how emotional design and purpose-driven messaging can reshape digital brand communication.
Artificial intelligence is rapidly transforming marketing, offering new ways for brands to connect with consumers across digital platforms. One of the most intriguing innovations? Virtual influencers: computer-generated characters that engage audiences on social media in lifelike ways.
These digital personalities are no longer just experimental novelties. With over 75% of U.S. users aged 18 to 24 following at least one virtual influencer, and 60% of marketers reporting collaborations with them in 2023, virtual influencers have become a serious tool in the marketing arsenal. Brands are using them not only to promote products but increasingly to support prosocial campaigns around sustainability, diversity, and social responsibility—values that resonate strongly with Gen Z.
But choosing the right virtual influencer and crafting the right message is far from straightforward. Should your campaign feature a human-like virtual influencer or a stylized anime-like one? Should the message focus on product features or social values? These decisions aren’t just aesthetic; they can significantly impact how consumers perceive your brand, whether they trust the message, and, ultimately, whether they’re willing to buy.
Despite the growing popularity of virtual influencers, marketers still lack clear guidance on how visual style and message framing interact to shape consumer responses. While psychological theories suggest that human-like influencers may build more trust, other design theories informed by biology argue that cute, non-human-like characters can trigger stronger emotional engagement. So which approach works best—and when?
How We Tested the Impact of Virtual Influencer Design and Messaging
To explore how virtual influencer design and message framing affect consumer responses, we conducted a large-scale online experiment using realistic social media content. The goal was to understand how different combinations of visual style and message type influence key outcomes like trust, purchase intention, and emotional engagement.
Participants were randomly shown one of four Instagram-style posts featuring a fictional virtual influencer named Ella, promoting a thermal mug from the fictional brand TREKMUG. In some posts, Ella appeared as a realistic, human-like woman; in others, she took on a stylized, anime-inspired look with large eyes and soft features—both common styles in real-world campaigns.
The messaging also varied: some posts focused on functional product benefits (e.g., insulation performance), while others emphasized prosocial values (e.g., environmental sustainability).
More than 3,200 U.S.-based participants took part in the study. After viewing the post, they completed a structured questionnaire measuring their reactions. Using established psychological scales to ensure reliability, we assessed purchase intention, perceived trustworthiness, and emotional perceptions like cuteness.
What Makes Virtual Influencers Effective
Our study revealed that design and message framing matter, and they can be strategically combined for optimal performance.
Anime-style influencers outperformed realistic, human-like ones in campaigns that focused on social values like sustainability. When the message was purely focused on functional product aspects, the difference disappeared— showing that emotional design works best when paired with meaningful purpose.
Messages emphasizing prosocial values consistently led to higher purchase interest than those focused on product features.
Interestingly, in the prosocial condition, anime-style influencers were also seen as more trustworthy than those with a human-like appearance. But trust alone didn’t explain their success; emotional impact played a key role.
The Power of Cuteness
To understand the strong performance of anime-style influencers, we looked at their visual traits. Their large eyes and soft features align with the baby schema, a concept from biology known to trigger feelings of warmth and care.
Participants rated anime-style influencers as significantly cuter, especially when paired with prosocial messages. This cuteness helped build trust and increased purchase interest, creating a powerful emotional response.
Designing Influence with Purpose
As virtual influencers become a fixture in digital marketing, the question is no longer whether to use them but how to use them effectively. Our research shows that emotional design is more impactful than realism, especially when campaigns aim to promote values like sustainability, fairness, or social responsibility.
Anime-style virtual influencers, with their soft features and expressive faces, don’t just attract attention; they build trust, evoke empathy, and amplify the emotional power of prosocial messages. Their visual warmth makes audiences more receptive to purpose-driven content, especially when paired with language that activates moral thinking and future-oriented values.
For marketers, this means rethinking design strategy. Virtual influencers offer a flexible way to embody brand values, tell emotionally resonant stories, and connect with audiences on a deeper level. In a crowded digital landscape, what stands out isn’t realism—it’s relatability, warmth, and emotional clarity.
Ultimately, designing influence with empathy isn’t just a creative choice but also a strategic advantage. Especially in cause-related marketing and corporate social responsibility initiatives, emotionally engaging virtual influencers can help brands inspire action, build loyalty, and shape a more meaningful consumer experience.
Authors
- Dr. Carolin Kaiser, Head of Artificial Intelligence, NIM, carolin.kaiser@nim.org
- Dr. Vladimir Manewitsch, Senior Researcher, NIM, vladimir.manewitsch@nim.org
- Linus Biedermann, Aalen University
- Prof. Dr. Anja Dieckmann, Alumni, anja.dieckmann@hs-aalen.de
Contact

![[Translate to English:] [Translate to English:]](/fileadmin/_processed_/1/2/csm_CK_NIM_01_58bed63910.png)







