Almost human? A case studies on the social media presence of virtual influencers
- Karina
- 3 days ago
- 10 min read
Generation Z and Generation Alpha are becoming increasingly codependent on technology and social media. As they spend unprecedented amounts of time online devoting hours to platforms like TikTok we are observing a profound behavioral shift. They are consuming rapid-fire content, becoming less traditionally literate, and living a significant portion of their lives through digital screens.
This deep digital immersion presents a fascinating predicament for modern marketing: If our target audience prefers interacting with screens over physical reality, do we even need real humans to market to them anymore?
It is incredibly tempting for executives to look at these trends and conclude that a flawlessly programmed, algorithm-driven virtual avatar is the ultimate marketing solution.
However, the data tells a completely different story.
I analyzed four major academic studies examining the deployment of artificial intelligence and virtual influencers in consumer markets [1, 2, 3, 4]. The research demonstrates that when it comes to marketing and selling to actual human beings, the requirement for an authentic emotional connection is undeniable. Even as technology adoption peaks, consumers actively rebel against artificial interactions that feel too deceptive.
In this breakdown, we will explore what these studies reveal about the human psyche.
We will uncover how consumers actually experience AI, why highly realistic avatars often trigger anger instead of trust, and how "perceived authenticity" ultimately dictates purchasing behavior.
Interestingly, buried within this data is one specific, highly counter-intuitive avatar design strategy that actually works across these digitally native generations to build deep trust. I will share exactly what that strategy is towards the end of this article.
But first, let me show you what the science actually says about why your hyper-realistic AI is currently failing to connect.
The Form vs. Behavioral Realism Mismatch
Marketers often push for the most visually stunning avatars, believing that a human-like face automatically builds trust. However, examining the theoretical boundaries of avatar marketing design reveals a critical flaw in this logic [1].
Research shows that an avatar's "Form Realism" (how human it looks) creates immediate, subconscious expectations for its "Behavioral Realism" (how intelligently it acts) [2]. As the study explicitly outlines, "As the form realism of an avatar increases, so do customers’ expectations for its behavioral realism" [1]. If a brand builds a bot that looks exactly like a human, the customer will instinctively expect it to converse, reason, and empathize exactly like a human.
When a highly realistic avatar provides simplistic, scripted answers, the illusion shatters. Customers experience a "negative disconfirmation of expectations," directly dropping their likelihood to purchase and severely damaging their overall satisfaction [2].
Consider the real-world example of the Swedish bank Nordnet, which launched a highly realistic avatar named Amelia. Because Amelia's programming only allowed for basic intelligence, she could not provide the complex stock purchasing advice her appearance implied she could handle [2]. The bank ultimately had to discontinue her because "Amelia's realistic anthropomorphic appearance might have led customers to develop high behavioral expectations, which Amelia could not deliver, giving rise to a negative disconfirmation" [1].
Similarly, IKEA eliminated its realistic avatar, Anna, after realizing her human-like appearance prompted complex customer questions that vastly exceeded her predetermined, scripted capabilities [1]. Her hyper-realistic face wrote a check her algorithmic brain could not cash.
Key Takeaways
An avatar's visual humanness (Form Realism) directly dictates the customer's expectation of its intelligence (Behavioral Realism) [1].
Highly realistic avatars that provide scripted, simplistic responses cause a negative disconfirmation, severely damaging the customer experience and purchase likelihood [1].
If a brand cannot afford advanced behavioral AI, it must lower the visual realism of the avatar (e.g., using a cartoonish design) to manage customer expectations effectively [1].
To prevent negative disconfirmation and lost sales, ensure your avatar's visual appearance never exceeds its actual cognitive capabilities.
The 4 Psychological Traps of Customer AI
Marketing technology vendors consistently sell AI based on data efficiency and targeting accuracy. However, understanding the sociological and psychological costs consumers experience when interacting with AI reveals a dangerous blind spot in this engineering mindset [2].
When brands optimize strictly for algorithmic efficiency without addressing human context, they actively destroy their brand equity [2]. Consumers do not view AI simply as a neutral tool; they experience it emotionally and sociologically. According to the research, when AI is deployed carelessly, customers fall into four distinct psychological traps [2]:
First, consumers feel exploited by data capture [2]. While marketers see data collection as a way to fuel personalization, consumers experience a black box of surveillance. Without transparent, granular control over their privacy, the perceived loss of control triggers hostile reactance and churn [2].
Second, consumers feel insulted by misclassification [3]. Algorithms constantly categorize users to serve targeted ads or recommendations. When an AI puts a consumer in the wrong demographic box or makes a biased assumption, the customer does not just see a software glitch - they feel profoundly misunderstood and morally outraged [2].
Third, consumers feel replaced by task delegation [3]. Automation is sold as a time-saver, but human beings derive self-worth from completing tasks and mastering their environment. When AI completely takes over a process without allowing the user to provide input or maintain autonomy, consumers feel useless and stripped of their self-efficacy [2].
Fourth, consumers feel alienated by fake social interactions [2]. Deploying hyper-realistic chatbots to scale customer service frequently backfires. Forcing a simulated, emotional connection with an entity that cannot actually feel empathy results in profound alienation, especially during sensitive or frustrating service interactions [2].
Key Takeaways
Focusing solely on the data efficiency of an AI tool ignores the heavy psychological toll the interaction takes on the end user [2].
Consumers interacting with AI face four specific psychological threats: feeling exploited, insulted, replaced, and alienated [2].
Stripping away a customer's autonomy, privacy, or sense of identity through automated algorithms immediately degrades trust and brand equity [2].
To prevent AI from sabotaging your customer journey, you must actively design against the four psychological traps by prioritizing consumer control, transparency, and human context over pure algorithmic efficiency.
Perceived Authenticity Drives Purchase Intention
Once a brand navigates the Uncanny Valley and aligns its avatar's behavioral realism, the ultimate goal remains driving revenue. To understand how to achieve this, we must examine the premise of measuring how young consumers' emotional connections to virtual influencers directly translate into buying behavior [3].
Many brands assume that if an avatar is visually stunning and posts frequently, it will automatically generate sales. However, the data reveals a much more nuanced reality.
The key finding from recent quantitative analysis demonstrates that emotional attachment to a virtual influencer is heavily predicted by "perceived authenticity" rather than just visual attractiveness or engagement frequency [3].
While visual beauty can initially capture attention, it is insufficient to drive action on its own. For aesthetic appeal to be effective in driving purchase intent, it must be paired with emotional meaning [3].
This brings us to the operational mechanism of avatar marketing. To translate passive views into actual revenue, an avatar must possess a consistent identity and emotionally relatable conduct [3]. Virtual influencers who demonstrate perceived authenticity—through cohesive storytelling, consistent character design, and reliable interactions—build the emotional trust necessary to influence consumer choices, especially for high-involvement products [3].
Ultimately, consumers do not buy from an avatar simply because it looks good or posts often. They buy because the avatar's curated identity feels authentic enough to trust.
Key Takeaways
Emotional attachment to a virtual influencer is driven primarily by perceived authenticity, not just visual attractiveness or high engagement metrics [3].
Visual beauty only influences purchase intent when it is connected to a deeper emotional meaning [3].
Brands must prioritize cohesive storytelling, consistent identity, and emotionally relatable conduct to convert virtual influencer audiences into paying customers [3].
To translate views into revenue, prioritize building your virtual influencer's perceived authenticity through consistent, emotionally relatable storytelling rather than relying solely on visual aesthetics.
The "Uncanny Valley" of Social Sentiment
When brands deploy an AI avatar, the instinct is almost always to make it as indistinguishable from a real human as possible. The assumption is that high visual realism leads to high relational trust.
To test this premise, researchers analyzed how human versus virtual influencers actually perform in publicly visible human networks [4]. The study compared the Instagram posting behaviors and user comment reactions across three distinct profiles: a real human lifestyle influencer (Marta Cygan), an anime-like, clearly computer-generated virtual influencer (Noonoouri), and a highly realistic, human-like virtual influencer (Lil Miquela) [4].
If the standard marketing assumption holds true, the human-like avatar should perform almost as well as the real human, while the cartoonish avatar falls behind.
The empirical data proved the exact opposite.
The hyper-realistic avatar (Lil Miquela) successfully generated massive vanity metrics, securing the highest number of likes and video views among the three [4]. A marketer looking only at a standard analytics dashboard would consider this a massive success. However, an analysis of the actual sentiment in the comments revealed a toxic underlying reality.
While the human-like avatar drove views, it also triggered a massive spike in negative emotions, anger, swearing, and doubting terminology [4]. The researchers noted that "the one with more human-like features elicited the least positive affect, with additional overtones of scepticism and dislike for her 'robotic' nature" [4].
This phenomenon is known as the "Uncanny Valley." As a non-human entity becomes more human-like, it eventually reaches a threshold where its slight imperfections make people deeply uncomfortable, resulting in a feeling of creepiness and psychological reactance. When an AI tries to "pass" as human, consumers feel deceived, and they aggressively disassociate from it.
The most fascinating contrast in the study came from the anime-like avatar (Noonoouri). Because it was overtly fake and made no attempt to deceive the audience into thinking it was a real human, it bypassed the Uncanny Valley entirely. In fact, the data showed that "users reacted with approximately equal amounts of positive affect terms towards [the anime-like avatar] and [the human influencer]" [4].
The lesson here is stark: Users will gladly accept and engage with virtual avatars, as long as those avatars are not pretending to be real [4]. Attempting to maximize "humaneness" does not provide a marketing advantage; it provides a liability.
Key Takeaways
Hyper-realistic virtual influencers drive high vanity metrics but trigger negative sentiment, anger, and psychological reactance [4].
Overtly fake, cartoonish virtual influencers bypass the Uncanny Valley and elicit positive emotional reactions on par with real human influencers [4].
Consumers willingly accept AI avatars as long as the technology does not attempt to deceive them by passing as human [4].
To protect your brand equity, design virtual influencers that openly embrace their artificial nature rather than attempting to flawlessly mimic human reality.
Practical Highlights & Key Steps
When looking at the landscape of consumer AI, there appears to be a stark paradox: attempting to make an AI look perfectly human triggers deep psychological backlash [4], yet emotional attachment and brand trust are the primary drivers of actual sales [3].
The breakthrough lies in realizing that visual realism is not the same as psychological authenticity.
Consumers do not demand a pixel-perfect human clone to make a buying decision; in fact, trying to pass off a virtual avatar as a real human is perceived as inherently deceptive and creepy [4]. True attachment is driven by "perceived authenticity"—the consistency of the character's identity, the quality of its social engagement, and its emotional relatability [3]. By shifting your focus from physical anatomy to behavioral trust, you can leverage virtual agents to scale your brand without threatening the customer's psychology.
Here is your step-by-step blueprint to implement this balance into your marketing funnel:
Step 1: Downgrade Form Realism
Do not build hyper-realistic human models to perform basic, low-complexity tasks [1]. If your digital assistant primarily answers customer FAQs, processes returns, or guides users through a simple search, use a stylized or obviously non-human design [1]. By lowering the visual "Form Realism," you immediately lower the user's subconscious expectations of "Behavioral Realism," preventing the disappointment and negative disconfirmation that destroys conversion rates [1].
Step 2: Embrace "Authentic Fakeness"
Stop attempting to hide your AI's digital origins. Instead, lean openly into its virtual nature [4]. Overtly computer-generated, stylized, or cartoonish avatars completely bypass the Uncanny Valley because they remove the psychological threat of deception [4]. This transparency establishes "authentic fakeness," which allows your audience to safely suspend their disbelief and fully engage with the playful, escapist, and entertainment value of your virtual content [4].
Step 3: Build Validation Loops
Protect consumer autonomy and self-efficacy by keeping your audience in the driver's seat [2]. If your AI makes recommendations or groups your users into specific classifications, do not make these decisions in a hidden black box [2]. Implement interactive UI elements that explicitly ask the customer to validate or edit the AI's inferences [2]. For example, prompt them with: "Our algorithm thinks you might like this—is that accurate, or should we update your taste profile?" Giving users the power to correct the machine turns an intrusive data-surveillance experience into an empowering partnership [2].
Step 4: Audit for Consistency
If you deploy a virtual influencer to represent your brand, prioritize the consistency of its narrative over its physical attractiveness [3]. Conduct regular audits of your avatar's character identity, backstory, and communication tone. Because young consumers' buying intentions are directly influenced by how "real" the character's conduct feels, your digital ambassador must display a highly stable persona [3]. Visual aesthetics can capture attention, but only a cohesive, relatable character identity will translate that attention into sustained brand trust and revenue [3].
Key Takeaways
The Visual-Cognitive Balance: An avatar's visual appearance must never write a cognitive check that its behavioral programming cannot cash [1].
Transparency as a Shield: Overt virtuality eliminates the threat of deception, establishing an "authentic fakeness" that is both safe and entertaining for the consumer [4].
Control Preserves Trust: Safeguard customer relationships by building feedback loops that allow consumers to validate classifications and modify automated outcomes [2]
To unlock the true power of AI marketing, build virtual ambassadors that are transparently digital in appearance, highly consistent in character, and relentlessly human in their respect for customer autonomy.
If you are ready to implement smart automation, optimize your AI marketing strategy, and engineer sustainable business impact and growth, here is how you can contact me to discuss your project.
Karina Gerszberg
Fractional CMO
An emerging theory of avatar marketing Authors: Miao, F., Kozlenkova, I. V., Wang, H., Xie, T., & Palmatier, R. W. (2022). Journal of Marketing, 86(1), 67-90. Link: https://journals.sagepub.com/doi/10.1177/0022242921996646
Consumers and artificial intelligence: An experiential perspective Authors: Puntoni, S., Reczek, R. W., Giesler, M., & Botti, S. (2021). Journal of Marketing, 85(1), 131-151. Link: https://journals.sagepub.com/doi/10.1177/0022242920953847
Quantitative analysis of virtual influencers: a computational approach to understanding young consumers' emotional connections and purchasing behavior Authors: Li, J. (2026). Data and Metadata. Link: https://dm.ageditor.ar/index.php/dm/article/view/799
Almost human? A comparative case study on the social media presence of virtual influencers Authors: Arsenyan, J., & Mirowska, A. (2021). International Journal of Human-Computer Studies, 155, 102694. Link: https://www.sciencedirect.com/science/article/abs/pii/S1071581921001129
Comments