011 min
The problem door-in-the-face addresses
A founder wants ten customers to join a six-month research panel. A product manager wants a customer to complete a twenty-minute survey. Each of them needs a yes to something that costs the other person time. Asked directly, many people say no.
The question behind the door-in-the-face technique is whether the order of requests changes the answer. If you ask for a lot first and then ask for less, do you get more yeses than if you had only asked for less?
The name describes the first step. The first request is expected to be refused, as if a door were closed on the person asking. The second request is the one the asker actually wants.
022 min
Where door-in-the-face comes from
Robert Cialdini and five co-authors published the first experiment in 1975, in the Journal of Personality and Social Psychology. Their paper is titled "Reciprocal concessions procedure for inducing compliance: The door-in-the-face technique."
The study used three groups. Group 1 was asked to volunteer to counsel juvenile delinquents for two hours a week for two years. The experimenter expected a refusal of a request that large. The experimenter then asked a smaller favour: to chaperone juvenile delinquents on a one-day trip to the zoo. Group 2 heard only the zoo request. Group 3 had the large request described to them, but was then asked only for the zoo trip.
Group 3 matters because it rules out a simple explanation. If merely hearing about a big commitment made the zoo trip sound easy, group 3 would have agreed as often as group 1. It agreed far less often.
50% of the participants in group 1 agreed to the small request, compared to 17% in group 2 and 25% in group 3.
Cialdini's team named the cause in the title of the paper: reciprocal concessions. The asker gives something up by moving from a large request to a small one, and the person asked feels they should give something up in return.
In 2021, Oliver Genschow and colleagues published a direct replication of the study in the same journal, 45 years after the original. Their compliance rates were similar to Cialdini's. Many classic findings in social psychology have not replicated, so this result is worth noting.
032 min
Why door-in-the-face works
Researchers agree that the effect exists. They do not agree on why it happens. Four explanations have been tested, and the evidence for each is partial.
| Explanation | What it claims | What a study found |
|---|---|---|
| Reciprocal concessions | The asker gave ground, so the person asked gives ground too. | The 2021 replication was designed as a critical test of this explanation and reported that its results support it. |
| Social responsibility | The person asked holds a personal rule that helping others matters. | Participants rated door-in-the-face scenarios as closer to helping than to bargaining. |
| Self-presentation | The person asked wants to look good to the asker. | In one study the technique worked better on friends than on strangers, the opposite of other studies. |
| Guilt reduction | Refusing the first request makes the person feel guilty, and agreeing to the second request relieves it. | In one study only people who were made to feel guilty and told the second request was equally helpful complied more. |
The studies do not give one winner. One study that compared reciprocal concessions with social responsibility found neither explanation sufficient on its own, and its authors suggested the two work together.
The practical reading is that the technique does not depend on a trick of wording. It depends on the person feeling that the situation has a social obligation in it. When that feeling is removed, the effect disappears.
042 min
Door-in-the-face in studies outside the laboratory
Several studies took the technique out of the original classroom setting.
A door-to-door donation request. Researchers asked residents for a large commitment (10 hours of volunteering over several weeks) or a moderate one (a $30 donation) and then asked for a small donation of any amount. Door-in-the-face produced more compliance overall than the small request alone. The increase was larger for strangers than for friends.
Cheese in the Austrian Alps. An experimenter selling cheese first offered two pounds for eight euros and then one pound for four euros. Every door-in-the-face condition produced more compliance than the control that only offered one pound. The version that stressed the concession ("two pounds is probably too much anyway") and the seller's credibility together produced the largest increase.
An internet fundraising page. A website asked visitors to spend several hours a week finding donors for children injured by landmines. After a refusal, visitors were sent to a page of links to charities. Those visitors clicked more links than visitors who had not been asked first. The study counted clicks, not donations.
A worksheet in a second-grade class. Children who refused a 100-question worksheet and were then asked to do 20 questions complied more often than children in the control group and the foot-in-the-door group. They also needed less help from an adult.
One caution applies to all four. They are reported through a secondary summary, and none of them is a product experiment. Treat them as evidence that the effect is not limited to one laboratory, not as a measured result for a pricing page.
052 min
Door-in-the-face in marketing and product work
The technique is easiest to see where a person makes two separate requests in one conversation.
A founder recruiting a design partner (illustrative). The founder asks a customer to join a monthly review call for a year. The customer says no. The founder then asks for one 30-minute interview. The decision that matters is the size of the first request: the 2021 replication compared an extreme request with a similarly small one, and the extreme request was the one that raised compliance.
A product manager asking for feedback (illustrative). The manager asks a customer to join a long diary study. After a refusal, the manager asks for a five-question survey. The second request must be a real reduction, and it must be the thing the manager would be glad to receive on its own.
The first request has to be one the person could plausibly be asked, even if they refuse. An absurd request may not read as a concession. This is a design judgment, not a finding from the studies.
In a written interface the two requests are separate screens or messages. A call to action that asks for a paid plan and then offers a free trial is a loose parallel, but none of the studies above tested that case, so treat it as a hypothesis to test.
Marketers often describe the large first request as an anchor. That description is a useful comparison for price talks, but it is not the explanation the experiments tested.
061 min
How to apply door-in-the-face honestly
Decide the request you actually want.
Write it down first. The technique does not work if you would be unhappy to hear yes to the first request.
Make the first request real but large.
It should be a request someone could reasonably accept, such as a year of calls instead of one.
Ask the second request straight away.
In the restaurant study, the effect appeared when the second request followed at once and not when the waiter waited three minutes.
Make the reduction visible.
Say plainly that you are asking for less. The cheese study found the largest effect when the seller stressed the concession.
Leave the refusal respected.
Accept the first no without argument. If you press, the second request becomes a continuation of pressure instead of a concession.
Measure it with a comparison. Send the small request alone to one group and the two-step version to another, then compare agreement rates, as the original study did.
071 min
When door-in-the-face fails
The second request is delayed. In the restaurant study, a waiter asked patrons if they wanted dessert and, after a refusal, asked about tea or coffee. Asking at once increased compliance. Waiting three minutes did not.
The person does not feel guilty or obliged. In a guilt study, only one of four conditions produced the effect: the one that made participants feel strongly guilty about refusing and told them the second request was equally helpful. In the other conditions, compliance was the same as in the control condition.
The first request is too small. The 2021 replication compared the extreme first request with a similarly small one and found only the extreme request raised compliance.
The person feels tricked. None of the sources measured this, so it is a reasoned limit and not a finding. A person who sees the sequence as a tactic may refuse the second request as well, and may trust the asker less afterward.
The compliance is verbal only. A study of a restaurant bill that a confederate asked participants to share found that door-in-the-face produced more actual payment than the control, not just verbal agreement. Even so, a promise is different from an action, so measure what people do.
081 min
Door-in-the-face versus foot-in-the-door
The nearest neighbour is the foot-in-the-door technique. Both are sequential requests, and the order of the requests is the one axis that separates them.
| Door-in-the-face | Foot-in-the-door | |
|---|---|---|
| First request | Large, expected to be refused | Small, expected to be accepted |
| Second request | Small, the real goal | Large, the real goal |
| Main mechanism proposed | Reciprocal concessions, guilt | Consistency with the earlier yes |
A meta-analysis of 22 studies that compared the two techniques found no significant difference in effectiveness. Because they work by different mechanisms, the right choice depends on whether a person is more likely to feel a social obligation after a refusal or to want to stay consistent with a first yes. See also commitment and consistency for the foot-in-the-door mechanism, and reciprocity for the norm behind concessions.
?6 questions
