People appear more attractive when seen in a group than when evaluated individually. The brain averages faces in a crowd, smoothing out individual imperfections. In UX, this means testimonials, team pages, and social proof clusters feel more persuasive when presented as a group rather than one at a time.
In 2014, psychologists Drew Walker and Edward Vul published a study with a deceptively simple finding: faces are rated as more attractive when seen in a group than when seen alone. They called it the “cheerleader effect.” The mechanism is straightforward -- when we look at a group, our visual system computes an ensemble average. Individual imperfections get smoothed out. The average face of a group is always more symmetrical, more proportional, and closer to the population mean than any individual face. And faces closer to the average are consistently rated as more attractive.
The effect is not limited to faces. It extends to any collection where the brain computes an ensemble impression before evaluating individual members. A row of testimonials feels more persuasive than any single testimonial. A team page with twelve headshots communicates more competence than any one person's profile. A logo wall of five partner brands creates more trust than any single logo. The group elevates each member by smoothing out individual weaknesses and creating a composite impression that is stronger than its parts.
In product design, the group attractiveness effect operates wherever multiple items are presented together. Understanding it means understanding why some social proof strategies work dramatically better than others -- and why the presentation format matters as much as the content.
“Individual faces seem more attractive when presented in a group because the visual system computes ensemble representations that are more average -- and more average faces are more attractive.”
— Walker & Vul, 2014
The same three testimonials feel fundamentally different depending on whether they are shown together or one at a time. In a group, each testimonial borrows credibility from the others. In isolation, each one has to stand entirely on its own merit -- and the weaker ones are exposed.
Notice that the 4-star review actually helps the group version. In isolation it feels underwhelming. In the group, it reads as honest -- a signal that the testimonials are real, not curated. The group attractiveness effect does not require every member to be perfect. It requires the average to be strong.
Startup “about us” pages face a fundamental challenge: how do you communicate competence when you are a small team nobody has heard of? The group attractiveness effect provides the answer -- present the team as a cohesive unit rather than a list of individuals. The group creates a composite impression of collective expertise that is stronger than any individual resume.
The group version is more persuasive not just because the individual presentations are better -- though they are -- but because the grid layout forces the brain to process the team as a unit first. The “30+ years combined” framing sets the ensemble impression before any individual is evaluated. By the time you read each person's card, the group halo has already shaped your expectations upward.
Partner and press logo walls are the most common application of group attractiveness in SaaS marketing. The logic is simple: if your product is used by Stripe, Shopify, and Notion, a visitor concludes your product must be good -- because those companies are good. The group halo transfers from the recognisable brands to yours. But the effect reverses when the group includes weak members.
The lesson is counterintuitive: a smaller group of strong members is always better than a larger group with weak ones mixed in. Five recognisable logos outperform ten logos where only three are recognisable. The group attractiveness effect rewards curation over quantity.
The group attractiveness effect is one of the simplest biases to apply ethically. Show your strongest social proof signals together. Present teams as cohesive units. Cluster testimonials rather than isolating them. The group does the work of elevating each member -- you just need to curate the group well enough that the ensemble average is high.
The danger is padding. Adding weak members to a group -- fake testimonials, unknown logos, filler team members -- does not just fail to help. It actively drags the group's perceived quality down. The brain computes the average, and a low-quality outlier pulls that average toward mediocrity.
Walker, D., & Vul, E. (2014). Hierarchical encoding makes individuals in a group seem more attractive. Psychological Science, 25(1), 230–235. — Ariely, D., & Levav, J. (2000). Sequential choice in group settings. Journal of Consumer Research, 27(2), 279–290.