Skip to content

Mindset

10 min left
BOOK SUMMARY · PERSUASION

Influence Summary: The Key Ideas, What Holds Up, and How to Use Them

Robert Cialdini's principles of persuasion in plain words, with the famous studies set beside later research.

Hacks Vitae artwork (AI-assisted)
The answer

Influence argues that people often say yes on a single cue, and that persuaders lean on six such cues; the 2021 edition adds a seventh, unity.

It suits anyone who wants to recognise the tactics used on them. Several techniques work in controlled tests, but the effects are often smaller than its stories suggest.

Why it mattersKnowing which tactics really move people helps you notice them when you are the one being asked.

Key facts
  1. ShortcutsPeople often respond to a single cue; Cialdini calls it "click, whirr".
  2. Door in the faceMoves spoken agreement; its effect on what people do is small.
  3. Low-ballReliable in a meta-analysis of published studies.
  4. Hotel towelsDisputed; a real but small effect is likelier, in our reading.
  5. ScarcityRaises how much people say they want something.
  6. UnityNew in 2021: favouring people who share an identity with us.

By Hacks Vitae Published Updated Last checked October 2026 10 min read 30 sources ↓

Start hereIn Langer's photocopier study, an empty reason after "because" worked almost as well as a real one for a five-page favour. For twenty pages, the empty reason did no better than none. So which of the book's techniques hold up?

About the book

Influence is by Robert B. Cialdini, Regents’ Professor Emeritus of Psychology and Marketing at Arizona State University. William Morrow first published it in 1984 as Influence: How and Why People Agree to Things. The 1993 edition carried the subtitle The Psychology of Persuasion. A textbook version, Influence: Science and Practice, has been published alongside it.

The 2021 New and Expanded edition adds a seventh principle, unity. It is written for general readers and mixes experiments with everyday examples. In the author’s own description of the 2021 edition, it shows how the principles work and how to defend yourself against unethical influence attempts.

We run on shortcuts

Cialdini’s starting point is that people often respond to a single cue instead of weighing all the information, much as some animals react to one trigger. He calls these automatic responses “click, whirr”. An early example in the book is a study in which a stranger asked to jump the queue at a photocopier.

Does it hold up? Only for a small favour, and repeat attempts disagree. In Ellen Langer’s 1978 photocopier study, asking to copy five pages worked 60% of the time with no reason. It worked 93% of the time with an empty reason (“because I have to make copies”) and 94% with a real one.

For twenty pages, the empty reason did no better than none (24% each), while the real reason got 42%. The small groups held about fifteen people each. The authors reported the large-favour gap only as a trend, in groups of about 25.

Two 1985 studies by Valerie Folkes did not replicate the result; a 2009 study by Key and colleagues did.

Try it: when you ask for something, give the real reason. It was the one version of the “because” effect that did not fall flat when the favour grew.

Reciprocity

People feel obliged to give back what they have received first: a gift, a favour, a concession. One tactic the book describes is what it calls the rejection-then-retreat technique, better known as the door in the face. Ask for something large, and when it is turned down, retreat to the smaller thing you wanted, which feels like a concession to be repaid.

Does it hold up? More people say yes; the effect on what they then do is smaller. Cialdini and colleagues’ 1975 experiments found the big-then-small sequence produced more compliance with the smaller favour, and a 2021 direct replication found compliance rates similar to the original’s.

A 2012 meta-analysis of 35 years of studies found a significant effect on verbal compliance (78 comparisons). It found no significant effect on behavioural compliance, meaning whether people actually did the thing (39 comparisons).

Its 2025 update by the same lead author, covering 50 years, found both effects significant but small. The figures were r = .14 for verbal compliance across 89 comparisons and r = .08 for behavioural compliance across 53.

Try it: if you agreed to something on the spot after turning down a bigger request, look at it again later, before you act on it.

Commitment and consistency

Once people have taken a stand, they tend to act in line with it. The book’s examples include the foot in the door, where agreeing to a small request makes a later, larger one easier to accept. Another is the low-ball, where a deal agreed at one price tends to survive a small increase in the cost.

Does the foot in the door hold up? It can work, but the effect is small and inconsistent. The classic 1966 study by Freedman and Fraser is the book’s example. A 1999 review by Jerry Burger argued that the processes involved can explain both successful demonstrations and studies in which the technique was ineffective or even lowered compliance.

A 2020 direct replication noted that effect sizes in meta-analyses tend to be low, and that the size of the original effect has never been obtained again. In its own tests the technique worked in Ukraine but not in Poland in 2003. It had no significant effect in either country in 2013.

Does the low-ball hold up? Yes, in published studies. A 2015 meta-analysis of published low-ball studies found the procedure “a reliable and effective method for increasing compliance”. It works best when people state their first agreement publicly and when the second cost is only slightly higher. It covered published studies only.

Try it: when extras appear after you have said yes to a deal, look again at the full price.

Social proof

When unsure what to do, people look at what others like them are doing. Cialdini and colleagues later tested this with hotel signs telling guests that most other guests reused their towels.

Does it hold up? Researchers disagree on how strong the evidence is. In the original 2008 study, co-written by Cialdini, the message about other guests beat the standard environmental message (44.1% against 35.1% reuse in one experiment).

Five later replication experiments that measured how many guests reused their towels each failed to reach significance. Among them, a 2014 German study found the norm message no more effective than the standard one. A separate 2014 study in two hotels, which counted towels used, found guests used significantly fewer towels when the message described what other guests in their own situation had done.

A 2016 Bayesian analysis that pooled all seven experiments found strong evidence for the effect. Critics argued the pooling was flawed, and that models allowing the studies to differ weaken the evidence. The authors agreed that analysing the studies separately suited this case better, and replied that averaging across models still gave fairly constant evidence in its favour.

Our reading: a real but small effect over the standard message is likelier, since the pooled evidence points that way and no single study was strong on its own.

Liking

People say yes more readily to those they like. The author’s own summary names three causes of liking: similarity, compliments, and working together towards shared goals.

In one 2004 paper, people who believed they shared a birthday, a first name or a fingerprint type with a requester were more likely to grant the request.

Authority

People tend to follow credible, knowledgeable experts, and outward signs of authority, such as diplomas on the wall and uniforms, can sway people too. The book discusses Stanley Milgram’s obedience experiments over several pages.

Does it hold up? The obedience held up; the explanation is argued over. In Stanley Milgram’s 1963 study, 26 of 40 participants gave what they believed was the highest shock.

Across the 21 versions of the experiment in which obedience meant going to the maximum shock, 43.6% of 740 people did, so how many obey depends heavily on the set-up. A 2006 partial replication, stopped at 150 volts for ethical reasons, found obedience only slightly lower than Milgram’s.

A 2015 study of Milgram’s archive argues that participants went on because they identified with what they took to be a worthwhile scientific project. That account is at odds with Milgram’s own explanation. A 2020 re-analysis found that in 18 of 23 versions of the experiment, those who fully believed the shocks were real gave lower shocks on average.

Try it: before following an expert, ask what they know about this particular question, and what they gain from your answer.

Scarcity

People want more of what they can have less of. The author’s own example is the rush to buy seats on Concorde after British Airways announced in 2003 that it would stop the twice-daily London to New York flight.

Does it hold up? Supported, mostly in what people say they want. In a 1975 experiment, cookies in scarce supply were rated more desirable than the same cookies in abundant supply.

A 2022 meta-analysis of 416 effect sizes from 131 studies starts from the finding that scarcity cues generally raise a product’s value and the intention to buy it. It maps which kinds of scarcity work best for which products. Both measure ratings and intentions rather than actual purchases.

Try it: when you see a countdown or “only 3 left”, ask whether you wanted the thing before you saw it was scarce.

Unity

New in the 2021 edition. In an article announcing it, Cialdini wrote that people are inclined to favour someone who shares a significant social identity, “someone they consider one of them”. Its power, he wrote, comes from shared identities such as family, nationality or religion rather than shared tastes.

What holds up, and where the book overstates

What the sources show for each idea

01 Getting someone to agree to a small request makes them more likely to agree to a bigger one later (the foot in the door)Credited to Robert Cialdini Partly It can work, but the effect is small and inconsistent across studies. Evidence ↓
02 Asking for something big first, then backing down to a smaller request, gets more people to say yes (the door in the face)Credited to Robert Cialdini Partly More people say yes; the effect on what they then do is smaller. Evidence ↓
03 Once people agree to a deal, they tend to stick with it even when the cost then rises a little (the low-ball)Credited to Robert Cialdini Holds up A meta-analysis of published studies found it reliable, best after a public first agreement. Evidence ↓
04 Telling hotel guests that most guests reuse their towels gets more of them to do itCredited to Goldstein, Cialdini and Griskevicius Still argued Five replications were not significant alone; a pooled analysis found strong evidence, which critics dispute. Evidence ↓
05 Thirty-eight neighbours watched Kitty Genovese being murdered in 1964 and did nothingCredited to The 1964 news story, as the book uses it Doesn't hold up A 2007 archive study found no evidence for 38 inactive witnesses; two people did call police. Evidence ↓
06 Adding "because" and any reason at all, even an empty one, gets people to agreeCredited to Langer's 1978 study, as the book uses it Partly Only for a small favour, and repeat attempts disagree. Evidence ↓
07 Most ordinary people will obey an authority figure who tells them to hurt someoneCredited to Milgram's studies, as the book uses them Partly Obedience held up but depends heavily on the set-up; why people obeyed is argued over. Evidence ↓
08 Things seem more valuable when they are scarceCredited to Robert Cialdini Holds up Supported, mostly in ratings and intentions rather than actual purchases. Evidence ↓

The broad picture holds: people do take mental shortcuts, and several of the techniques work in controlled tests. But the size of the effects is often smaller than the book’s stories suggest, and one famous story, the 38 silent witnesses, is not supported by the evidence.

The 38 silent witnesses

The book uses the 1964 killing of Catherine (Kitty) Genovese in Queens to explain why a crowd of onlookers can fail to help. The textbook edition quotes the news story’s line that “38 respectable, law-abiding citizens in Queens watched a killer”.

A 2007 study of the archive material found no evidence that there were 38 witnesses, that they saw the murder, or that they did nothing.

In 2016 the Times itself called the portrayal of 38 fully aware and unresponsive witnesses erroneous. Some neighbours did ignore her cries, but the original article grossly exaggerated how many witnesses there were and what they saw. Afterward two people did call the police.

The broader finding is real but narrower than the story. A 2011 meta-analysis found that each person is less likely to help when others are present, an effect weaker in dangerous emergencies.

Who it’s for, and who can skip it

It suits readers who want a lively, story-led tour of persuasion, and anyone who wants to recognise the tactics used on them in shops, sales calls and fundraising.

Readers who want effect sizes and replication status will need to look past it to the later research. Readers who already know the six principles from courses or summaries will find little new in the core ideas.

Quick answers

What are Cialdini's principles of persuasion?

Reciprocity, commitment and consistency, social proof, liking, authority and scarcity. The 2021 edition adds a seventh, unity.

Does the door-in-the-face technique work?

It moves spoken agreement more than action. A 2025 review of 50 years of studies found both effects significant but small: r = .14 for verbal compliance and r = .08 for behavioural compliance.

Did 38 people really watch Kitty Genovese's murder and do nothing?

The evidence does not support it. A 2007 study of the archive found no evidence that there were 38 witnesses, that they saw the murder, or that they did nothing, and in 2016 the Times called its portrayal erroneous.

Did the Milgram obedience experiment hold up?

The obedience did: a 2006 partial replication, stopped at 150 volts, found obedience only slightly lower than Milgram's. Why people obeyed is still argued over.

Use this

The lesson

Most of these cues work quietly, so the defence is a pause at the moment of yes.

The rule

Would I still say yes if nobody had pushed me to say it now?

Quick quiz

Three questions, thirty seconds

Nothing is scored or saved. Recalling it once now is what makes it last.

1. In Langer's study, how did an empty reason do when the favour was large?

Answer: No better than no reason.Read it again ↑

2. What did a 2007 study of the archive find about the 38 witnesses to Kitty Genovese's murder?

Answer: No evidence there were 38 who saw it and did nothing.Read it again ↑

3. In the 21 versions of Milgram's experiment where obeying meant going to the maximum shock, what share of people did?

Answer: 43.6%.Read it again ↑

How we checked this · Last checked October 2026
  • What we read: short search-inside passages of the 1984 and 1993 printings and of the 2009 textbook edition, the author's own pages on his principles and books, and his 2021 article announcing unity, which is how the unity principle is described here.
  • For the studies we read the full texts of the 1978 photocopier study, the 1985 replication, the 2016 towel re-analysis and its 2017 reply, and the 2016 New York Times obituary (archived copy); for the others, the abstracts.
  • The verdict on the hotel towels is our reading of a dispute between researchers; both sides are on the page.

How we check every article →

Sources

Primary written at the time, by a witness, or the original text Scholarly edited or peer-reviewed work Popular press, blogs and general websites