What is social proof in marketing?
Cialdini's Influence says social proof works under three conditions. Field studies from Beijing restaurants, hotel bathrooms, a national park and a news aggregator show what each one does for a marketer.
Answer to the question
What is social proof in marketing?
Social proof in marketing means showing evidence of what other people have chosen, such as review totals, sales counts or ratings, so that uncertain buyers follow them. Cialdini names three conditions under which it works best: uncertainty, numerous others and similarity to the buyer.
Key takeaways
- Cai, Chen and Fang (2009) found that a plaque listing a restaurant's five most popular dishes raised demand for them by 13 to 20 per cent, while a plaque naming sample dishes without a ranking had no significant effect.
- Goldstein, Cialdini and Griskevicius (2008) reported towel reuse in 44.1 per cent of hotel rooms with a descriptive-norm card against 35.1 per cent for the standard appeal; Bohner and Schlüter (2014) found norm messages beat no message but not the standard appeal.
- Salganik, Dodds and Watts (2006) ran 14,341 participants through parallel music markets, and where counts were visible the same song could finish near the top in one world and near the bottom in another.
- Muchnik, Aral and Taylor (2013) found an unearned first up-vote raised the likelihood of a positive rating by 32 per cent, while an unearned down-vote was corrected away.
- Cialdini's account of the Petrified Forest signs gives wood removal by 2.95 per cent with no sign, 7.92 per cent with a three-people sign and 1.67 per cent with a lone-person sign.
In a Beijing field experiment, showing diners the five most popular dishes raised demand for them, and naming dishes without a ranking did not.
That gap is most of what social proof in marketing is: showing people what others have chosen, so that an uncertain buyer can borrow a decision. It works when the information is true, comes from people like the buyer and describes something worth copying.
Do plaques and towel cards beat a plain argument?
Hongbin Cai, Yuyu Chen and Hanming Fang ran that experiment on restaurant tables in Beijing, with plaques assigned at random. The plaque with the ranking carried the lift, and the plaque that only named sample dishes did not, so the lift came from the popularity claim and not from the mere sight of a dish name. The effect was modestly stronger among infrequent customers, and dining satisfaction rose with the top-five plaque. The diners who knew least leaned on the crowd most, and the crowd was telling them something true.
That is the first of the three conditions Cialdini gives in Influence: uncertainty. The second is the many, meaning evidence from numerous others. The third is similarity, evidence from people who resemble us.
Social proof works best when the evidence comes from people similar to us.
Hotel bathrooms supply the test of similarity. In a 2008 field experiment, Noah Goldstein, Cialdini and Vladas Griskevicius left towel cards in hotel rooms. A card saying that most guests reuse their towels produced reuse in 44.1 per cent of rooms, against 35.1 per cent for the standard environmental appeal. In a second experiment, a message citing guests who had stayed in the same room reached 49.3 per cent. Cialdini calls the pull of such near neighbours peer-suasion.
The sign itself replicated.
Bohner and Schlüter ran a version in two German hotels, with 928 observations, stating that 75 per cent of guests reused their towels. Descriptive-norm messages raised reuse above a no-message baseline but did not beat the standard environmental appeal, and the effect of closeness to the reference group was inconsistent across their two studies. A social proof message can be expected to beat silence. Beating a good plain argument is a harder claim, and the evidence for it is thinner than the textbook version.
What happens when the count describes something nobody wants?
A different test of the idea began in a national park. Cialdini and colleagues tested signs against wood removal at Petrified Forest National Park, and the paper appeared in Social Influence in 2006. By his account in Influence, 2.95 per cent of visitors removed wood with no sign. A sign picturing three people taking wood, which told visitors that taking wood was common, raised it to 7.92 per cent. A sign picturing one person taking wood cut it to 1.67 per cent. The paper's abstract reaches the same direction: messages about what others do were most likely to increase wood removal, and messages about what others disapprove of were most likely to suppress it.
Cialdini calls the error the Big Mistake, which is to normalise a behaviour in the act of lamenting how common it is. Social proof states what is normal whatever the sentence around it intends. His later advice, set out under future social proof, is to describe an honest trend rather than a current majority when the current majority is on the wrong side.
A household test shows both effects at once. In a 2008 field experiment, Nolan and colleagues found that messages about what other people do produced the largest change in energy use, while participants rated that same information as the least motivating. Cialdini gives the figures in Influence. Four door-hanger messages went to homes, and the one telling residents that most of their fellow community members try to conserve energy at home produced 3.5 times the saving of any other message, measured on the meters. Homeowners surveyed beforehand had expected it to be the least effective.
Social proof can be manufactured by a practitioner rather than arising on its own.
The plaque, the towel card and the door hanger used social proof to tell people what many similar others did, and it was something worth copying. The forest sign gave them the same signal about something nobody wanted.
All four worked as designed.
Does a visible count decide which option wins?
A count can change which option wins, which is a separate effect from changing how much of one option sells. Matthew Salganik, Peter Dodds and Duncan Watts built an artificial music market in 2006 in which 14,341 participants listened to previously unknown songs and chose which to download. Some could see how often each song had been downloaded and some could not.
The market ran as parallel worlds, and where the counts were visible the same song could finish near the top in one world and near the bottom in another. The best songs rarely did poorly and the worst rarely did well, but between those ends the order depended on social influence. [Why the best song does not win](https://thewhyweproject.com/articles/why-the-best-song-does-not-win) follows that study in full.
What does a fake count actually buy a marketer?
A visible count is easy to fake, and a field experiment shows what a fake buys. Lev Muchnik, Sinan Aral and Sean Taylor worked with a social news aggregator whose administrators up-voted, down-voted or left alone the first comment on more than 100,000 posts. An unearned up-vote raised the likelihood of a positive rating by 32 per cent and final ratings by 25 per cent on average. An unearned down-vote was corrected away, because other readers pushed back. Unearned approval compounded and unearned disapproval did not.
Before publishing a count, a marketer has two questions to settle: who the number describes, and whether it is real. Its size comes a distant third. A review total, a bestseller badge and a customer number are all claims about what other people did. Where they did it independently and resemble the reader, the claim carries evidence. Where the crowd was staged, or the number describes a behaviour the seller would rather not see repeated, it still moves people, and that is the problem. Honest tallies and staged ones sit under the same label on the page, and nobody prices them differently.
Cialdini draws his line at the same place. It is exploitation, he writes, when the lever is not a natural feature of the situation but is contrived by the practitioner.
Common questions
What is social proof in marketing?
Social proof in marketing means showing evidence of what other people have chosen, such as review totals, sales counts, ratings or testimonials, so that uncertain buyers follow them. Cialdini names uncertainty, numbers and similarity as the three conditions under which it works best.
What are examples of social proof in marketing?
A restaurant plaque listing the five most popular dishes raised demand for them by 13 to 20 per cent in Beijing. A hotel card saying that most guests reuse towels lifted reuse to 44.1 per cent from 35.1 per cent. Review counts, bestseller badges and customer numbers belong to the same family.
Does social proof always increase sales?
No. In two German hotels, norm messages beat a no-message baseline but did not beat the standard environmental appeal. A count that describes a behaviour sellers want to stop can raise it, as the Petrified Forest signs did.
Is fake social proof effective?
In a randomised experiment on a news aggregator, an unearned first up-vote raised final ratings by 25 per cent on average, while an unearned down-vote was corrected away. Cialdini defines exploitation as contriving the lever, so effective and legitimate are separate questions.
Sources
Cialdini, R. B. (2021). Influence, New and Expanded: The Psychology of Persuasion. New York: Harper Business. Seed book; vault hub Books/Influence.
Cai, H., Chen, Y., & Fang, H. (2009). “Observational learning: Evidence from a randomized natural field experiment.” American Economic Review, 99(3), 864-882. doi:10.1257/aer.99.3.864 https://doi.org/10.1257/aer.99.3.864
Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). “A room with a viewpoint: Using social norms to motivate environmental conservation in hotels.” Journal of Consumer Research, 35(3), 472-482. doi:10.1086/586910 https://doi.org/10.1086/586910
Bohner, G., & Schlüter, L. E. (2014). “A room with a viewpoint revisited: Descriptive norms and hotel guests' towel reuse behavior.” PLOS ONE, 9(8), e104086. doi:10.1371/journal.pone.0104086 https://doi.org/10.1371/journal.pone.0104086
Salganik, M. J., Dodds, P. S., & Watts, D. J. (2006). “Experimental study of inequality and unpredictability in an artificial cultural market.” Science, 311(5762), 854-856. doi:10.1126/science.1121066 https://doi.org/10.1126/science.1121066
Muchnik, L., Aral, S., & Taylor, S. J. (2013). “Social influence bias: A randomized experiment.” Science, 341(6146), 647-651. doi:10.1126/science.1240466 https://doi.org/10.1126/science.1240466
Cialdini, R. B., Demaine, L. J., Sagarin, B. J., Barrett, D. W., Rhoads, K., & Winter, P. L. (2006). “Managing social norms for persuasive impact.” Social Influence, 1(1), 3-15. doi:10.1080/15534510500181459 https://doi.org/10.1080/15534510500181459
Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2008). “Normative social influence is underdetected.” Personality and Social Psychology Bulletin, 34(7), 913-923. doi:10.1177/0146167208316691 https://doi.org/10.1177/0146167208316691