Should Your Brand Hire a Virtual Influencer?

When the meal-kit delivery company HelloFresh wanted to promote its latest line of healthful menu items, it made an increasingly popular choice: It hired an Instagram lifestyle influencer. Jenna Kutcher is a Minnesota-based mother of two who has more than a million followers, and as part of a 21-day challenge, she and 15 other influencers hired by the brand posted recipes and photos of meals they made using HelloFresh, each tagged #RefreshWithHelloFresh. The campaign produced 461 influencer posts and generated 5.5 million impressions, with 20% of the influencers’ followers mentioning HelloFresh on Instagram—a clear success.

Results like that helped influencers earn some $21 billion in 2023. But such partnerships are not without potential pitfalls. Influencers’ credibility is built on trust, which can prove delicate. Sometimes a post is perceived as inauthentic, or the influencer exhibits off-platform behavior that’s not aligned with the image or values of the brand being promoted. And as influencer marketing grows, so do examples of promotional relationships that caused disillusionment and regret.

In 2017, for example, Adidas Originals created an Instagram ad featuring Kendall Jenner as its brand ambassador. Detractors argued that there was nothing “original” about Jenner: “[Has she] really faced it all? What could a bourgeois like [her] have possibly faced?” In late 2023 China’s “Lipstick King” Li Jiaqi, a pioneer among fashion and cosmetic influencers with 76 million followers, publicly lost his temper with one who complained about the price of eyeliner, excoriating the commenter for not working hard enough to grow his income. The incident prompted a backlash and a tearful public apology by Li. When Chriselle Lim, a fashion and beauty influencer, collaborated with Volvo on posts promoting the carmaker’s eco-friendly product line, critics highlighted the disconnect between the green promotion and Lim’s high-consumption, materialistic lifestyle.

Those incidents raise a question: Can you hire an influencer whose careful behavior and reliable image limit the risk of undercutting the promotional message? Yes, you can.

Lil Miquela is a social media influencer who has 3.5 million followers on TikTok and another 2.7 million on Instagram. Since her debut, in 2016, she has earned an average of $2 million a year posting content on behalf of brands including Dior, Calvin Klein, and BMW. Lil Miquela is far less likely than traditional influencers to say or do something scandalous, because she’s a computer-generated image, programmed and controlled by a marketing company.

Although human influencers still vastly outnumber virtual influencers, the latter are becoming more common. According to one survey, 52% of U.S. social media users already follow a virtual influencer, and that percentage is higher globally. Brands including Prada, Cartier, Disney, Puma, Nike, and Tiffany use virtual influencers to promote their products. In this article we draw on our academic research and studies by others to examine how brands should choose between a human influencer and a virtual one. We also offer insights that brands can use when developing their influencer marketing strategy.

Measuring the Pros and Cons

To understand the advantages and disadvantages of human influencers and virtual influencers, we focus on five factors: engagement, reach, diversity, reputation risk, and cost.

Engagement.

To learn how consumers react to virtual as opposed to human influencers, we conducted research on the way followers engage with posts from each type. We collected marketing content posted by 551 human influencers and 13 virtual influencers, all sponsored by 112 brands. Most of the brands are in the fashion and beauty segments, with a handful representing technology, travel and lifestyle, health care, or other industries. The content, which included more than one million posts, was put online between June 2014 and December 2020. Some of it was sponsored, and some was organic. We measured how much followers liked and commented on various types of influencer content.

Prior studies have shown that when a virtual influencer (whether anime-like or humanlike) creates organic Instagram posts, they tend to receive more likes and more positive emojis than do the posts of human influencers. Research by Chen Lou of Nanyang Technological University and colleagues identified what drives consumers to engage more with virtual influencers, including the novelty of interacting with computer-generated imagery (CGI) and the different aesthetic it represents. In addition, consumers exposed to virtual influencers’ posts show greater willingness to share their favorable experiences with others.

In our research we focused on sponsored posts—the kind marketers arrange and pay for. When we compared paid posts by virtual influencers with their organic content, we found that followers showed 13.3% more engagement with paid than with nonpaid posts, whereas sponsored posts by human influencers garnered 2.1% less engagement, on average, compared with the same influencers’ organic content. Among the industries we studied, the fashion and beauty sector was especially hospitable to sponsored posts by virtual influencers—and also showed more resistance to sponsored posts by human influencers: Virtual influencers’ paid posts in that sector drew 16.3% more engagement than did their organic content, on average, while human influencers’ paid posts drew 2.3% less engagement than their unpaid postings.

The engagement data shows a clear upside to using virtual influencers.

Reach.

Brands choose to use big-name celebrities as influencers for a reason: Celebrities have lots of followers, which increases exposure for their products. In December 2020 the average number of followers for human influencers in our study was about 2.8 million, whereas the average number for virtual influencers stood at about 1.1 million.

In academic research about influencers, the number of followers and consumer engagement have an inverted U-shape relationship—that is, influencers with relatively higher or lower numbers of followers create less engagement than influencers with moderate followings do. That suggests that brands needn’t necessarily be put off by virtual influencers’ smaller audiences. Our research found that when a brand is not widely known, the number of followers an influencer has doesn’t play a big role in how much engagement a piece of sponsored content sparks. Among the fashion brands in our sample, for instance, we found that lesser-known brands such as Misbhv and Moschino achieved greater engagement in response to sponsored virtual-influencer posts than did more-established brands such as Burberry, Chanel, Dior, and Louis Vuitton.

Diversity.

Modern brands want to be inclusive, and that requires partnering with influencers who represent a variety of demographics. Human influencers are not a particularly diverse group. Marketers in certain segments, such as fashion and beauty, say the majority of influencers in their field are female and white. When we used an AI program to try to detect the racial background of the 551 human influencers in our study, it estimated that 68% were white, 11% were Latino/Hispanic, 10% were Black, 8% were Asian, 1% were Middle Eastern, and 1% were Indian. In theory, brands could try to hire from those shallow pools, but they may face other complications: The influencers may already have deals with competing brands, or be promoting so many other brands that they’re not taking new clients, or be so geographically distant that it is logistically challenging to work with them, or have an aesthetic that doesn’t align with what the brand wants, or have accents or speak languages that may be difficult for the brand’s target market to understand. Thus brands have begun looking to virtual influencers to achieve diversity.

When marketers are creating an influencer with computer graphics, they have no constraints on race, gender, or other characteristics. As an example of how extreme the diversity can get, one virtual influencer, named Blu, is fashioned as an extraterrestrial creature orbiting Earth in his spaceship, the Xanadu. Companies show imagination in building their own influencers by seeking the optimal combination of demographics, behavioral traits, and personality characteristics.

Ralph & Russo, a British fashion house, successfully used a virtual influencer to launch its 2020–2021 couture collection. The company designed Hauli, a tall Black virtual model. Her name comes from the Swahili words for strength and power. The campaign showed images of her posing at the Taj Mahal, the Great Wall, and other wonders of the world—places where conducting a photo shoot with a human influencer would be difficult (and expensive). The combination of an African influencer and a global context proved successful. The promotion achieved 19.4 million views worldwide, and the company estimated the value of the media exposure at $65.1 million.

Reputation risk.

Human influencers periodically become embroiled in scandal or notoriety. Brands try to manage this risk, but it is ever present. Virtual influencers, devoid of the autonomy that typically accompanies human behavior, mitigate it.

Nars, a French cosmetics and skin care company, turned to virtual influencers after such a scandal. In 2018 the YouTube beauty influencer Manny Mua and others were featured in a group photo giving the finger to and denigrating a rival beauty influencer named Jeffree Star. The post created a backlash that the influencer media labeled Dramageddon. Mua lost a large portion of his followers, and Nars severed ties with him. The following year Nars launched a trio of virtual influencers—Maxine, Chelsea, and Sissi—who behaved the way their computer handlers programmed them to. Not surprisingly, their content has been drama-free.

Cost.

Put simply, human influencers are more expensive than virtual influencers. For instance, a human influencer with a million-plus followers might charge a brand more than $250,000 per post. The company that created Lil Miquela, currently the most popular virtual influencer, charges only $9,000. Because virtual influencers are cheaper—and don’t have travel expenses that must be covered—brands can hire larger numbers of them.

Making the Choice

The concept of virtual influencers is just a few years old, so academic research is only beginning to understand how and why consumers react to this form of marketing. Nonetheless, our research and close observation of the field’s evolution suggest four principles to help guide brands in their choice.

Recognize consumers’ desire for novelty.

When it comes to endorsers and influencers, some brands find value in stability and longevity. For example, Tiger Woods’s recently dissolved relationship with Nike lasted 27 years. In the social media era, however, many consumers crave something new and different. Brands that rely on familiar celebrities risk being perceived as stale or uninteresting. People scroll social media quickly, and it takes something unusual to make them pause. Indeed, research has shown that one reason consumers follow virtual influencers is that they are unexpected and previously unknown. Although betting on a virtual influencer may require a leap of faith, marketers should realize that human influencers may be perceived as overexposed or past their prime.

Look at the data.

Our research involved studying more than a million pieces of content over six years and doing statistical analysis. That level of research may not be feasible for every brand, but a fundamental analysis of social media data is both achievable and essential. Are your competitors’ brands using virtual influencers? If so, how do consumers appear to be engaging with that content versus posts by human influencers, and how is that changing? Are brands in your industry using more virtual influencers over time? Engagement data is public and easily accessible on Instagram and other platforms. Brand managers should collect and study the data relevant to their category.

Embrace a portfolio approach.

Brands typically do not rely on either human or virtual influencers, or on any individual within either category. Instead they choose a mix of influencers. In a recent study we looked at whether human influencers are being replaced as brands find success with virtual influencers. We discovered that brands that embrace virtual influencers continue to use human influencers as well, but they typically switch to different ones.

Experiment, measure, and learn.

Because virtual influencers represent a lower-cost marketing opportunity, brands can treat the choice to hire them as an experiment. They can deploy a virtual influencer or two and then closely watch engagement and results. They should calculate the return on investment and compare it with alternatives, including other virtual influencers, human influencers, and other kinds of marketing. The influencer industry is mature enough to have standard KPIs and metrics for judging performance, and brands can embrace this methodology as they evaluate decisions about influencer marketing.

. . .

As consumers spend less time on traditional media and more time on social media, influencer marketing is likely to become even more important for brands. Because it remains fairly new, fewer rules and established practices exist, which creates opportunities for innovation. Our research suggests that virtual influencers offer distinct advantages over traditional influencers. The success brands are having with them should inspire companies to be open to additional innovations as they embrace social media marketing.