Only some commitment and consistency techniques survive being named aloud

There is an honest way to help somebody keep a decision they have made, and it has one requirement: the decision has to already exist before you start building around it.

What to take away

• Helping a person hold to their own choice is legitimate. Manufacturing a choice from a small step they did not know was a step is not.
• Every honest technique here can be handed to the person as a written instruction and still work.
• If you are on the receiving end, the correction is to treat the small step as a small step rather than as a position.

Techniques that survive being explained

Ask for the decision, plainly, once. The clearest version of this whole area is simply asking. A person who says yes to a stated proposal has made a decision, and everything after that is support rather than persuasion.

Write down what was agreed, and send it. A shared record turns a conversation into a commitment that both sides can see. It also lets the person notice if what you wrote is bigger than what they said, which is precisely why sending it is the honest move.

Help them build their own device. Automatic transfers, booked slots, a standing appointment, a deadline told to someone who will ask. The person chooses the constraint and knows what it is for, which is the difference between a commitment device and a trap.

Make the first step genuinely small and genuinely optional. A trial that is easy to leave is fine. The test is whether leaving is as easy as arriving, and if it is not, the smallness was doing something other than reducing risk.

Ask what would make them change their mind. Doing this before the commitment protects both of you. It gives the person an exit they defined themselves, and it stops the arrangement turning into something they have to defend rather than something they want.

Give a way out, out loud. "If this stops working for you, tell me and we will stop." Said at the start, it costs nothing and removes the thing that makes people dig in. It also, reliably, increases how long they stay, which is inconvenient for anyone who wanted a story about how honesty is expensive.

Techniques that collapse when named

Technique	Why it works	What is never said
The ladder of small steps	Each step is too small to refuse	"None of the steps was the request"
Getting it in writing early	Written positions are hard to abandon	"I wanted it written before you were sure"
The public declaration	Witnesses hold people in place	"The audience is the mechanism"
Changing the terms after agreement	The decision is already made	"The first price was not the price"
Framing the exit as inconsistency	Nobody wants to be unreliable	"I made leaving feel like a character flaw"

The last row is worth separating out, because it does lasting damage. Telling someone leaving would contradict who they are turns a practical decision into an identity question.

Identity questions are harder to reason about. It is effective, and it is the clearest case on this page of influence that depends entirely on the person not seeing what is happening.

The fourth row has a specific commercial form: a benefit offered at the start, with an ongoing obligation that becomes visible later. The Federal Trade Commission's Negative Option Rule sets out what sellers must make clear in that arrangement. The staff report Bringing Dark Patterns to Light catalogs the wider family of designs that make leaving harder than joining.

If you are helping somebody follow through

Three rules cover most of it.

• Make the commitment specific and small enough to keep. Vague commitments fail and then teach the person that they are somebody who fails.
• Separate the goal from the identity. "I said I would run three times this week" recovers from a missed week; "I am a runner" does not, and the collapse is worse.
• Build the review in. A commitment with a scheduled date for asking whether it still makes sense is one the person can hold without defending, which is the whole point.

Each of these can be said aloud to the person it is being done to, which is the standard used across this site and set out under social influence basics.

When it is being used on you

The move that handles most cases is to name the step. "I am happy to do this small thing, and I am not agreeing to the larger one." Said once, in a friendly voice, at the time.

That sentence works because the technique needs the small step to be read as a position. Naming it as a small step removes exactly that, and it costs nothing with somebody acting honestly.

Whether your future self should be bound by what you say now, and how much, is genuinely contested, and the Stanford Encyclopedia entry on promises sets out why. The mechanism is described under commitment and consistency, the field guide for the engineered versions under manipulation resistance, and the way witnesses harden a stated position under normative influence.

Common questions

Is asking for a small yes always a tactic?

No. Starting small is often good practice and reduces risk for both sides. It becomes a tactic when the small step was designed to make the larger one hard to refuse and the person was not told.

Should I get agreements in writing?

Yes, and send them to the other party rather than filing them. A written record that only one side has seen is doing something different from a shared one.

Does a public commitment help people keep promises to themselves?

It seems to help some people and to make others worse, particularly when the announcement itself feels like progress. Try it once, on something small, and keep the result rather than the theory.

What if somebody wants to back out of a commitment to me?

Let them, and say plainly what it costs. Holding somebody in place with the argument that leaving would be inconsistent gets you a person who is present and no longer working with you.