Cialdini's 7 principles of persuasion: What They Are and Where They Fail

Scott Watson, Founder, Summit Consulting and Training
The UK's first Certified Influence Professional, personally certified by Dr Robert Cialdini.

Published July 2026

Most articles about Cialdini's principles will tell you what the seven are. This one also tells you where each of them fails. Which, after more than a decade studying this work and a fair few years teaching it to rooms full of sceptical managers, I've come to think is the more useful half.

One point of accuracy first, because it's stated wrongly nearly everywhere, including by people selling training in it. Robert Cialdini did not identify seven principles. He identified six, in Influence in 1984: reciprocity, commitment and consistency, social proof, authority, liking and scarcity. The seventh, unity, arrived thirty-two years later in Pre-Suasion, and was folded into the expanded edition of Influence in 2021. His own account is that it had been sitting in his data the whole time, he simply hadn't seen it.

What makes his work unusual is where it was done. Most social psychology of that era happened in laboratories with undergraduates. Cialdini's tradition was field-based: hotels, restaurants, doorsteps, car parks, real people spending real money. That's a large part of why these findings have held up better than their contemporaries. Though the effects are more modest than the popular retellings suggest, and I'll say so where that matters.

His central concern throughout has been ethical influence: the same seven levers can help someone reach a decision they'll be glad of, or manoeuvre them into one they'll regret. The difference usually comes down to whether you understand the failure modes. So each section below covers what the principle is, what the evidence actually shows, what it looks like in a working environment, where it backfires, and one thing you can try this week.

1. Reciprocity

What it is

We feel obliged to repay what we've been given. It is one of the most reliable rules in human social behaviour, and it operates whether or not we asked for the thing in the first place (which is exactly why it can be abused).

What the research shows

The clearest demonstration I use in training rooms comes from a study of restaurant tipping by Strohmetz and colleagues. Servers brought a sweet to the table with the bill. One sweet lifted tips a little. Two sweets lifted them considerably more. But the biggest increase came from a third condition: the server left one sweet, started to walk away, then turned back and offered a second... saying, in effect, "and because you've been so pleasant, here's another one for you." Same sweets. Very different outcome. What changed was that the gift became personal and unexpected.

What it looks like at work

Most organisations give constantly and get nothing back for it, because the giving is systematised to the point of invisibility. The automated welcome pack, the standard onboarding email, the generic newsletter. None of these trigger reciprocity, because nobody experiences them as a gift. They're just process.

What does work is the small, specific, clearly-chosen-for-you thing. A colleague forwards an article with two lines on why it made them think of your problem. A supplier sends the one page of their methodology that solves your particular bottleneck, before you've signed anything. A manager notices someone is stretched and takes one task off them without being asked.

Where it backfires

Three ways, and I see all of them.

The first is giving something the other person doesn't actually value. Obligation only arises if the gift registers as a gift. A branded notebook creates nothing.

The second is making the transaction visible. The moment someone recognises that the favour was a lever, reciprocity doesn't just fail, it reverses. People feel handled, and they withdraw. Cialdini's own guidance is that the recipient can defuse the whole mechanism simply by reclassifying the gift as a sales tactic.

The third is timing. The sense of obligation decays. If you give value in January and make your ask in June, you have banked goodwill but you have not banked reciprocity.

Try this

This week, give one prospect or colleague something genuinely useful, specific to their situation, with no request attached and no data capture. Then notice how the next conversation opens.

2. Commitment and Consistency

What it is

Once we've said or done something, we work hard to stay consistent with it. Consistency is socially prized - we distrust people who lurch about, so the pull is strong, and it operates on us long after we've forgotten what we originally committed to.

What the research shows

The foundational study is Freedman and Fraser's work in a Californian suburb. Researchers asked homeowners to put a large, deliberately unattractive road-safety billboard on their front lawn. Most refused, which is what you'd expect.

But a second group had been approached two weeks earlier with a far smaller request: display a three-inch window sticker supporting safe driving. Almost everyone had said yes to that. When the same group was later asked about the billboard, agreement rose dramatically.

Nothing had changed about the billboard. What had changed was how those homeowners now saw themselves as the sort of person who supports road safety.

What it looks like at work

This is why the sequence of asks matters more than the strength of any single one. "Would a fifteen-minute conversation about your priorities be useful?" is a different proposition from "shall we book the programme?", and one leads to the other far more reliably than a cold approach to the second.

It also explains why written commitments outperform verbal ones, and public commitments outperform private ones. If you want a change initiative to stick, having people write down what they'll do differently - in their own words, in front of colleagues does more than any amount of communication from the top.

Where it backfires

Two ways, and the second is the one leaders should worry about.

First, consistency delivers stated agreement much more reliably than it delivers behaviour. A meta-analysis covering thirty-five years of sequential-request research found a clear effect on verbal compliance and essentially none on behavioural compliance. People say yes and then don't act. If your programme evaluation rests on what participants agreed to in the room, you are measuring the wrong thing.

Second. And this is the uncomfortable one, the same drive that produces customer loyalty produces escalation of commitment. The manager who publicly championed a failing project is precisely the person least able to stop it, because stopping it means being inconsistent with a position they've already defended. Every organisation has a project that ran two years longer than it should have for exactly this reason.

Consistency is not a virtue in itself. It's a bias, and it's as capable of locking people into bad decisions as good ones.

Try this

Look at your current sales or change process and find the smallest genuine yes you could ask for before the real one. Then look at your last three stalled projects and ask who publicly committed to them, and whether that's why they're still running.

3. Social Proof

What it is

When we're uncertain, we look at what people like us are doing and treat it as evidence. It's an efficient shortcut most of the time. It's also the principle most easily faked, which is why it needs the most careful handling.

What the research shows

The best-known field test is Goldstein, Cialdini and Griskevicius's hotel towel study. Guests were given one of several cards asking them to reuse their towels. A standard environmental appeal produced a certain level of reuse. A card noting that the majority of guests reuse their towels did better. But the strongest result came from a card saying that the majority of guests who had stayed in that same room reuse their towels.

The lesson isn't that social proof works — everyone knows that. It's that proximity beats volume. "Most people like you, in your situation" outperforms "thousands of customers worldwide," even though the second number is far larger.

What it looks like at work

If you're a charity fundraiser, "23 similar-sized regional charities have run this" is more persuasive to your trustees than "used by over 10,000 organisations." If you're introducing a new process, evidence that the team next door has already adopted it will move people further than a case study from a FTSE 100 firm.

This is also why vague testimonials are almost worthless. "Great service, would recommend" carries no information about who the person was or what changed. Name, role, situation, specific result — that's what converts, because it lets the reader recognise themselves.

Where it backfires

This is the principle with the most dangerous failure mode, and it's rarely taught.

Descriptive norms don't only pull people up. They also pull people down. In a field experiment on household energy use, Schultz, Cialdini and colleagues told households how their consumption compared with their neighbours'. Heavy users cut back, as intended. But light users — the ones already doing the right thing — increased their consumption towards the average. The researchers called it the boomerang effect. Adding a signal of social approval for low use removed it.

Apply that to your own organisation and the implication is uncomfortable. Tell your team that the average salesperson makes twelve calls a week, and your best performers now have permission to do twelve.

The related trap is negative social proof: announcing that too many people are doing the wrong thing. In a study at the Petrified Forest National Park, signage highlighting how many visitors had removed petrified wood was associated with more theft than a control condition. The message was intended as a rebuke. What it actually communicated was that stealing is normal here.

If you are ever tempted to say "too many staff are still not completing their timesheets," this is the research that should stop you.

Try this

Take one piece of social proof you currently use and make it narrower rather than bigger. Swap the largest number you have for the closest match to the person you're speaking to.

4. Authority

What it is

We defer to people who appear to know what they're talking about. This is usually sensible — expertise is real, and checking everything ourselves is impossible — but the deference attaches to the signals of expertise, not to expertise itself, and the two can be separated.

What the research shows

The example I use most is an estate agency that changed nothing except how calls were transferred. Instead of "I'll put you through to Sandra," the receptionist added a sentence about why Sandra was the right person - her years in lettings, her specialism. Appointments and signed contracts both rose.

No new qualifications. No change in service. The expertise already existed; it simply hadn't been communicated.

There's a related finding worth knowing, because it runs against instinct. Authority rests on two things — expertise and trustworthiness — and most people over-invest in the first. Openly admitting a genuine weakness before presenting your strength tends to increase credibility rather than reduce it. The classic commercial examples are brands that led with their drawback: the mouthwash that acknowledged its unpleasant taste, the car hire firm that admitted to being second in the market.

What it looks like at work

Have someone else introduce you. Third-party credibility does work that self-promotion cannot, and a colleague can say things about you that would sound absurd in your own mouth.

Put your real credentials where people encounter them before they meet you — the email signature, the LinkedIn banner, the page they read before the call. Authority signals only work if they're seen in advance.

And if you're pitching, name a genuine limitation of your offer early. "This won't help you if X" costs you nothing with the wrong buyer and buys you considerable trust with the right one.

Where it backfires

Authority produces compliance whether or not the authority is right. That's the finding, and it should sit uneasily with anyone in a senior role.

The aviation industry worked this out the hard way. Analysis of cockpit accidents found cases where junior officers noticed a captain's error and either said nothing or hedged so softly that the warning didn't land. The industry's response — crew resource management — was essentially a programme to make authority safe to challenge. Most organisations have never done the equivalent.

The practical implication for leaders is inverted from the usual training message. If you hold authority, your problem is not projecting more of it. Your problem is that people are already agreeing with you more than your reasoning deserves, and you will be the last to find out.

Using this when you have no authority

The situation I'm asked about most in training rooms isn't selling. It's this: someone needs a colleague in another department to change how they work, and has no authority to make them.

That's where these principles earn their keep, because authority is the one lever you don't have. Unity and liking establish that you're on the same side rather than making demands from outside. Reciprocity means arriving having already helped rather than only ever asking. Commitment works better than pressure, because a small agreement someone made voluntarily outlasts one you extracted.

Influencing stakeholders sideways is a different skill from influencing a team downwards, and most management training never separates the two.

Try this

Ask your team, individually and privately, for one thing they think you've got wrong. If nobody offers anything, that's your answer.

5. Liking

What it is

We say yes to people we like. The main drivers are similarity, genuine compliments, cooperation towards a shared goal, and simple familiarity through repeated contact.

What the research shows

The most useful study for modern working life looked at negotiations conducted by email. One group went straight to business. The other spent a short period beforehand on an unrelated personal conversation — background, interests, nothing to do with the deal. The groups that talked first reached agreement far more often; the groups that went straight to business hit impasse at a markedly higher rate.

Five minutes of apparently irrelevant conversation changed the outcome of the negotiation. Anyone running remote or hybrid teams should read that twice.

There's also a finding about compliments that's worth reporting honestly, because it's slightly unflattering. Flattery tends to have a positive effect on how we feel about the source even when we consciously recognise it as insincere and know the person has something to gain. We discount it intellectually and are influenced by it anyway.

What it looks like at work

Sixty seconds of preparation before a call — finding one genuine, specific thing about the person's professional situation to open with — changes the temperature of the conversation more than any amount of polish in your pitch.

Cooperation matters more than most people realise. Liking is built faster by working alongside someone towards a shared outcome than by socialising with them. If you want a difficult stakeholder onside, find something small you can genuinely build together.

Where it backfires

Manufactured similarity is detected far more often than the person manufacturing it believes. Claiming to support the same football team, or discovering a suspiciously convenient shared interest, doesn't read as rapport. It reads as technique, and the recovery from that is slow.

The more serious cost, though, falls on the person being influenced rather than the influencer. Liking corrupts judgement. It shapes who gets hired, which supplier gets renewed, whose performance review is generous, and whose idea gets airtime in the meeting. That isn't persuasion working, it's decision-making failing.

A useful discipline in any significant decision: ask yourself whether you'd reach the same conclusion if you didn't like the person. If the answer isn't an easy yes, slow the decision down.

Try this

Before your next three meetings, spend a minute finding something specific and genuine to open with. Separately, review your last significant supplier or hiring decision and ask honestly how much liking was doing.

6.Scarcity

What it is

Things become more attractive as they become less available. Two forces drive it: we dislike losing options, and reduced availability is a rough proxy for value.

What the research shows

Worchel and colleagues gave people biscuits from a jar and asked them to rate them. Some saw a jar containing ten. Others saw a jar containing two. The biscuits from the near-empty jar were rated more desirable, despite being identical.

The most interesting condition was a third one. Some participants watched a jar of ten get replaced with a jar of two, apparently because other people had taken them. Those biscuits were rated highest of all.

Scarcity works. Scarcity that has just arrived, driven by other people's demand, works harder.

What it looks like at work

The honest version is straightforward: state your real constraints, specifically. "We take six new clients a quarter and two places remain" is scarcity. So is a genuine intake date, a cohort that runs once a year, or a diary that actually is full until September.

Loss framing matters too, and it's underused internally. "Here's what we stand to lose if we don't act" reliably outperforms "here's what we'd gain," which is why it deserves careful handling rather than enthusiastic adoption.

Where it backfires

Fake scarcity is the fastest way to destroy a reputation you spent years building. If your website has said "only two places left" every week since March, your customers have noticed, and they've quietly reclassified everything else you tell them as marketing.

Less obviously, scarcity can trigger resistance rather than compliance. When people feel a freedom is being removed rather than a genuine constraint being described, the reaction is often reactance — they push back, and they push back harder than they would have if you'd said nothing.

And inside organisations, manufactured urgency is corrosive. When everything is urgent, nothing is prioritised, decisions get worse, and people burn out. Leaders who use artificial deadlines to drive pace usually get pace for a quarter and cynicism thereafter.

Try this

Audit your marketing for any urgency claim you couldn't defend if a customer asked. Remove it. Then find one genuine constraint you've been too polite to mention and state it plainly, once.

7. Unity

What it is

The seventh principle, and the newest. Unity isn't about being similar to someone, that's Liking. It's about shared identity: the sense that you and they belong to the same category. Family, region, profession, cause, generation, a shared experience nobody outside it fully understands.

It's worth being accurate about the history here. Cialdini identified six principles in Influence in 1984. Unity arrived thirty-two years later, in Pre-Suasion, and was folded into the expanded edition of Influence in 2021. His own account is that it had been sitting in his data all along.

What the research shows

The clearest practical demonstration is also the simplest. When you ask someone for their opinion, they step back and evaluate you. When you ask for their advice, they step towards you and think alongside you. In tests of a new restaurant concept, consumers asked for advice responded considerably more positively than those asked for their opinion or their expectations — same concept, same question, one word different.

Advice creates a merged perspective. Opinion creates a separated one.

What it looks like at work

Describe who you serve by identity, not job title. "HR managers" is a market segment. "People who went into HR because they believe work should be better than this" is an identity, and the right people recognise themselves in it instantly.

Change one word in your next stakeholder conversation: ask for their advice on the problem rather than their view of your proposal. It's a small shift and the difference in how the conversation runs is disproportionate.

And watch your pronouns. "We" meaning your company is marketing. "We" meaning you and the client together is Unity.

Where it backfires

Unity has the sharpest edge of the seven, because every in-group implies an out-group.

The same force that makes a team cohesive makes it hostile to the department next door. Strong internal identity is why sales and operations end up unable to talk to each other, why the London office and the regional office develop mutual contempt, and why merged organisations stay two organisations for a decade. If you deliberately strengthen unity within a team, you have also, without meaning to, strengthened the boundary around it.

Cialdini has been explicit that unity helps explain political tribalism and the effectiveness of inflammatory headlines. That should give anyone pause before deploying it casually in marketing.

The other failure is claiming membership you haven't earned. Unity asserted from outside the group reads as appropriation, and it costs more trust than it could ever have gained.

Try this

Rewrite one headline to describe your ideal customer by who they are rather than what they do. Then identify the group your team currently defines itself against, and consider what that's costing you.

Conclusion

Understanding and applying Dr Robert Cialdini's seven principles can significantly enhance your ability to influence others ethically and effectively. Reciprocity, commitment and consistency, social proof, authority, liking, scarcity and unity give you a way to navigate almost any communication situation with more skill and less guesswork.

It's worth being honest about the scale of them, though, because the influence industry rarely is. These are tendencies rather than levers. When interventions of this kind are tested at scale, the effects tend to land in single digits rather than the transformations you'll hear promised at a sales seminar. Several of the studies here date from the 1960s and 70s with modest samples; they've held up better than most social psychology of that era, largely because Cialdini worked in the field rather than the laboratory, but the original numbers are noisier than the retellings suggest. That doesn't make them useless. A reliable few percent, applied consistently across every conversation your organisation has, compounds into something real. It just isn't magic.

The more important point is that ethical considerations aren't a caveat bolted onto this material, they're built into how it works. Every principle here has a failure mode. Reciprocity reverses when the gift is recognised as a lever. Consistency locks managers into projects they should have stopped. Social proof drags your strongest performers towards the average. Authority silences the person who might have caught your mistake. Liking quietly corrupts hiring decisions. Fake scarcity devalues everything else you say. Unity builds a wall around your team.

Notice what those have in common. In nearly every case, the damage lands on the person doing the influencing rather than the person being influenced — usually later, and usually invisibly. Which is why using these principles responsibly isn't just the decent thing to do. It's the only version that holds up over time, and it's how you end up with the positive relationships and genuine trust that make everything else easier.

So the skill worth building isn't deploying the seven. Most people manage that after an afternoon. It's knowing which one fits the situation in front of you, and recognising the moment when using it would cost more than the yes is worth. Get that right and everybody does genuinely win. Not as a nice sentiment, but because it's the only outcome that lasts.

To find out more about Summit Influence and Persuasion training courses for leaders, sales teams, marketers and customer relations teams, get in touch.

Scott Watson is the UK's first Certified Influence Professional personally certified by Dr Robert Cialdini to deliver Cialdini Insitute training and consulting programmes.