Show individuals, not numbers

Value is often communicated through numbers—statistics, reach, and magnitudes. But do large numbers actually move people? The key question is: How do people respond to statistical versus individual information? Which format is more persuasive, and what does the evidence tell us?

Studies

Only child defies statistics

Deborah Small and George Loewenstein conducted a groundbreaking experiment in 2003 that transformed our understanding of charitable giving. They presented 161 Carnegie Mellon University students with two different donation appeals for Save the Children. Half saw a photo of Rokia, a 7-year-old girl from Mali, accompanied by her personal story of hunger and thirst. The other half read dry statistics about millions of starving children in Africa. The individual case generated average donations of $2.38 per person, while the statistics yielded only $1.14. Even more striking: when both approaches were combined, donations dropped to $1.43—the numbers undermined the compassion evoked by Rokia's story.

Calculating Kills Compassion

In 2007, Small, Loewenstein, and Slovic sought to understand why statistics diminish compassion. They asked 93 students to either solve mathematical problems or evaluate emotional words before presenting them with a donation appeal. The only difference: five minutes of analytical versus emotional thinking. The results were striking: The 'calculators' donated only $1.26 instead of $2.34 for an identifiable child named Rokia—a 46% decrease. Notably, donations for statistical victims increased slightly. Even brief analytical thinking was enough to significantly weaken the emotional response to a concrete individual's plight.

Psychic numbing with large numbers

In 2007, Paul Slovic uncovered a disturbing phenomenon he termed "psychic numbing." Across several experiments, participants paradoxically placed greater value on saving a single life than on saving 1,000 lives. In a simulated famine scenario, people donated less money to help 11 million affected individuals than they did for "only" 3 million. The explanation: the larger the number, the more abstract the suffering becomes to our brains. A single child with a name and face activates our evolutionarily developed compassion system, whereas millions register merely as meaningless statistics.

The Taxi Problem

Daniel Kahneman and Amos Tversky conducted an experiment at Hebrew University in 1972 that became a classic in judgment research. Participants received the following information: In a city, 85% of taxis are green and 15% are blue. A taxi was involved in a nighttime accident. A witness identified it as blue. Tests showed the witness correctly identifies colors in 80% of cases. Question: How likely is it that the taxi was actually blue? Most participants answered 80%—they relied entirely on the witness testimony. The correct answer according to Bayes' theorem: only 41%. The striking finding: The clear base rate of 85% green taxis was virtually ignored. A specific description (the witness testimony) completely overshadowed reliable statistics.

The Mechanic and the Lawyer

Maya Bar-Hillel conducted experiments at Hebrew University in 1980 that demonstrated how even irrelevant descriptions displace base rates. Participants were told: In a group of 100 people, there are 70 engineers and 30 lawyers. Then a person was described: "Dick is 30, married, without children, very capable and highly motivated. He will be successful. His colleagues like him." Participants were asked to estimate: How likely is Dick to be an engineer? Most said 50%—the description was neutral, after all. The statistical reality: 70%. In a variant without any description, participants correctly used the base rate of 70%. The remarkable finding: An information-free description caused people to completely discard clear statistical information. Irrelevant details overrode relevant numbers.

Steve the Librarian

In 1973, Kahneman and Tversky conducted one of the most famous experiments on human judgment. They presented 85 Stanford students with a personality description: "Steve is shy, withdrawn, helpful but has little interest in people. He loves order and details." Then they asked the crucial question: Is Steve more likely to be a librarian or a farmer? Over 80% chose librarian, despite the fact that there are 20 times more farmers than librarians in the USA. The vivid personality description caused them to completely ignore statistical reality. Specific details overrode the base rate—a cognitive error that occurs systematically even among doctors and judges.

Natural frequencies help

In 1995, Gigerenzer and Hoffrage discovered a remarkable way to help even experts make better decisions. They presented 48 German physicians with an identical diagnostic task using two different formulations. Group A received percentages: "1% prevalence, 80% sensitivity, 10% false-positive rate." Group B received natural frequencies: "Out of 1,000 women, 10 have the disease, 8 of them test positive, and 99 healthy ones test falsely positive." The results were dramatic: only 10% of Group A calculated the probability correctly, compared to 46% of Group B. Yet the tasks were mathematically identical. Concrete numbers instead of abstract percentages quadrupled diagnostic accuracy.

Principle

Which principle for Customer Experience Design can be derived from this? People respond more powerfully to concrete individuals than to statistical masses—a principle with fundamental implications for customer experience design. While the brain processes abstract numbers and data rationally, personal stories and individual experiences activate emotional centers and drive stronger motivation and action. This effect proves particularly powerful for emotional topics and social causes, though it may be less relevant for purely functional product decisions. Companies must ensure that individual stories are both authentic and representative to maintain credibility. The following guidelines demonstrate how to implement this principle in practice.

Guidelines

Show impact through individual stories

When a company creates social impact, it should be presented through individual stories. Instead of stating "10,000 trees planted," say "We planted this forest—Maria and her family now live here." The following examples illustrate this guideline:

• TOMS Shoes: Each pair of shoes shows the child who receives a pair – with name and face. The impact is a person, not a number.
• Charity: Water: Donors receive GPS coordinates of 'their' well and updates about the people who use it. The connection is personal.

Customer Stories Instead of Testimonials

Tell concrete customer stories instead of listing features: Present a specific protagonist—including their name, age, profession, and situation—facing a problem. Show how your product solved it and what concrete results they achieved. A detailed story is more convincing than claiming 10,000 anonymous satisfied customers. The following examples illustrate this guideline:

• Airbnb: The 'Host Stories' feature presents complete narratives: Who is the host? What inspired them to start hosting? What experiences have they had? These stories convey emotion—and prove more persuasive than star ratings alone.
• Patagonia: 'Worn Wear' tells the stories of products and their owners: a jacket that survived three expeditions, shorts worn across three generations. These narratives convey quality without mentioning a single product feature.

Make customers visible

Show the people behind your brand – both your customers and your team members with real faces. Replace anonymous data points with personalized thank-yous; instead of '500 experts,' say 'This is Anna, who has been developing our products for 8 years.' Real people work here and buy from you, not faceless numbers. The following examples illustrate this guideline:

• Starbucks: The name on the cup is more than logistics. It signals: 'You're not customer #47, you're Sarah.' This minimal personalization activates connection.
• Duolingo: 'You've been with us for 30 days now!' – This message shows: We're tracking your progress. We see you. That motivates more than abstract streak numbers.

Replace statistics with customer stories

Instead of stating '95% customer satisfaction,' share the story of one specific customer: their challenge, your solution, and the result. Let numbers support the narrative, but never replace it. A detailed case study with a real name, context, and emotional journey proves far more persuasive than any statistic. The brain remembers people, not percentages.

Communicating risks through concrete scenarios

When communicating risks, don't rely on probabilities. Instead of saying "The failure probability is 0.1%," describe a concrete scenario: "Imagine the server goes down during your product launch—that means..." People can visualize scenarios but struggle to process abstract probabilities. Concrete worst-case scenarios activate caution more effectively than statistics.

Demonstrate quality through specific examples

Quality promises gain credibility through specific examples rather than metrics. Instead of claiming "99.7% defect-free production," demonstrate a concrete quality feature: "Every seam is inspected three times before..." Describe the process, identify the person responsible for inspection, and explain the reasoning behind it. Specificity builds trust because it demonstrates that someone understands the details intimately. Abstract percentages, by contrast, could easily be fabricated.

Replace success rates with typical results

When communicating success, describing 'what customers typically achieve' works better than stating 'X% reach their goal'. Describe the typical outcome concretely: 'Most of our customers reduce their process costs by 15-20% within three months.' This is tangible and credible. An abstract success rate of 87% remains meaningless. People need a concrete picture of the outcome, not a probability.

Always show overall rating

When displaying individual reviews, always show the overall context: average rating, number of reviews, and distribution. A single 1-star review appears devastating without context. However, when paired with "4.7 average from 1,247 reviews," it becomes properly contextualized. The following examples illustrate this guideline:

• Amazon-Review-Header: Above each individual rating: The overall rating and distribution graph - context always visible.
• App-Store-Bewertungen: Distribution bars (how many 5-star, 4-star, etc.) provide an immediate overview.

Visualize distribution

Display the rating distribution as a graph rather than only the average. A 4.0 average could represent all 4-star ratings or an equal split of 5-star and 3-star ratings. The distribution reveals the consistency of user experiences. The following examples illustrate this guideline:

• Booking.com: Horizontal bars for each star rating show the distribution at a glance.
• Indeed-Arbeitgeberbewertungen: Distribution by categories (work-life balance, salary, etc.) - more differentiated than an average.

Contextualizing negative reviews

Place negative reviews in context—"This is 1 of 247 reviews. 94% are rated 4-5 stars." Without explicit context, a negative review appears representative. With context, it's recognized for what it typically is: an isolated case. The following examples illustrate this guideline:

• Tripadvisor: In addition to critical reviews: 'X out of Y guests recommend this hotel.' - Context/classification.
• Software-Reviews: Note: This review refers to version 1.0. Current version is 3.2.' - temporal context.