How to Ask for Advice and Not Get Agreement

Ask for the objection, not the advice. People default to endorsement because disagreeing with you is socially expensive, and because you almost always present a decision you have already made, which leaves them nothing to do except approve it. Asking for honest advice is a procedure, and it starts before you describe the plan.

Key Takeaways:

• “Be honest with me” asks for a trait when what you need is a task, which is why it changes so little
• Endorsement is the cheap answer: agreeing takes one sentence and no risk, objecting takes five and a possible argument
• “What would have to be true for this to be a mistake” asks for a mechanism rather than a verdict, and mechanisms are easier to hand over
• Describing a decision you have already made turns your listener from advisor into witness, and they will behave accordingly
• Advice that arrives before you have finished asking was fired at the shape of the question, not at your situation

Why People Agree With You Even When They Disagree

Worth saying before any of it: none of what follows was measured by anybody. It is what turns up when you pay attention to how requests for advice get answered, and the only reason to believe a line of it is that you can test it the next time you ask someone something.

Agreement is the low-cost answer, and it is cheap in a specific way. It costs one sentence and zero relationship risk. An objection costs a paragraph, a possible argument, and the decent chance of being the person who called it wrong right before it worked out. Nobody is being a coward here. They are running arithmetic anybody would run.

The second half of the problem belongs to you. A great many requests for advice arrive after the decision, wearing a question as a disguise. You have seen the apartment, you like the apartment, and what comes out of your mouth is “what do you think of this apartment?” Your listener can hear the answer sitting inside the question. Telling you the kitchen is a problem now means telling you that you are wrong about something you have visibly already chosen, which is a much bigger favor than you meant to ask for.

Harder to catch is the version that feels like real consultation. You lay out the plan in detail, including the risks, and you sound admirably balanced. But almost all of that description goes on why the plan makes sense and one line goes on what could go wrong, and the ratio is the message. You have handed over a verdict and asked for a stamp.

If somebody transcribed your setup and read it back, would they be able to tell which answer you wanted? They can. That is why they gave it to you.

What “Be Honest With Me” Actually Asks For

It asks for a character trait, and the person was already being honest. They honestly do not want to spend Thursday evening fighting you about your apartment. What you meant was “please do the expensive work of finding the flaw,” and that is a task rather than a disposition.

There is a second effect, and it runs opposite to the intention. Framing the request as a demand for brutal honesty raises the stakes of the reply, so the objection they might have tossed off casually now has to justify the drama of the setup. Small doubts get filtered out. Small doubts are most of the useful material.

Replace the plea with an assignment.

Asking for Honest Advice: the Four Sentences That Change the Answer

The sentences below do the work because each one lowers the social cost of telling you something you do not want to hear; they belong to the same family as the deep questions worth asking people in general, tuned for a single decision. None of them require your friend to be braver than they are.

“I have not decided yet, and what is still open is X.” Say the specific undecided part, not a general claim of open-mindedness. “I have not decided” on its own is what everyone says on the way to doing the thing. “What is still open is March versus not at all” tells them exactly where their input can land.

“What would have to be true for this to be a mistake?” This is the strongest of the four. It asks for a mechanism instead of a verdict, and mechanisms are far easier to produce than judgments. It also hands over cover: they are describing a hypothetical world, not calling you wrong in your kitchen.

“Here is the strongest case against it that I already know about.” Handing them your own best objection does two jobs at once — it proves you can hear one without flinching, and it stops them spending their whole turn on the obvious risk you have already priced in. Now they have to find something you missed, which is the thing you actually came for.

“Who would tell me not to do this, and what would they say?” Asking somebody to voice an absent person’s objection is dramatically cheaper than asking them to own one. Your careful sister, your old boss, the friend who thinks you move too fast. Some you may not have spoken to in a year, which is its own reopening problem for old friends. People will say things in that frame that they would never say straight.

Order does more work than wording. Situation first. Your own leaning last, or never.

Try one swap this week: replace “what do you think?” with “what would have to be true for this to be a mistake?” and watch the pause before the answer get longer. The pause is the point. It is the same reason the “how was your day” question gets a one-word reply: a wide, cheap prompt earns a wide, cheap answer.

How to Read the Answer You Get

Speed is the first tell. An answer that arrives before you have finished the last sentence was fired at the shape of the question, not at your situation. Actual consideration is slow, and it is faintly uncomfortable to sit through, which is why people rush to fill it.

The second tell is whether they ask you anything at all. Somebody forming a real opinion needs facts they do not have, usually the money and what happens if it fails in month two, and they go looking for them. An endorsement needs nothing. It can be delivered by a person who was not listening.

Third, does the answer name a cost? “Sounds great” names nothing. “You would be paying rent in two cities for four months” names a cost, and it is useful whether or not the person turns out to be right, because now there is something to check.

Genuine agreement exists. The difference is that it can tell you what would change its mind, so make that the follow-up to every yes you receive: what would move you off that? If nothing would, you did not collect an opinion. You collected a mood.

Where This Backfires

Some people disagree by reflex, and their objection is worth exactly as little as an automatic yes. The friend who has never once said a plan sounded good is not providing a signal, because a detector that fires at everything carries no information. Weight what they say by how specific it is, not by how negative it is.

The worse case is asking somebody who has a stake in the answer. Your partner cannot be a neutral advisor on whether you take the job three states away, and asking them to be one is a way of getting a distorted answer and then being irritated at them for it. Ask them what they want instead. Different question, and one they can answer honestly.

There is also a ceiling on how much of this an ordinary conversation will carry. Running a full adversarial briefing about which contractor to use turns a chat into a hearing, and people quietly stop wanting to be asked things. Save the four sentences for decisions that are expensive to reverse.

Different people for different jobs, then. One for the mechanism, one for the emotional read, a job closer to listening without jumping to a fix, and nobody for both, because the second job contaminates the first.

FAQ

How do I get someone to be honest with me about my plan? Give them a task instead of asking for a trait. “What would have to be true for this to be a mistake?” asks for a mechanism, which is easier to produce and much cheaper socially than telling you that you are wrong. Add which part of the decision is genuinely still open, and mean it.

Why does everyone tell me my idea is great? Usually because you are presenting a decision rather than a question, and because agreeing costs one sentence while objecting costs an argument. Try describing the situation without your own leaning attached and see whether the responses change. If they still agree, ask what would change their mind.

Should I tell people I want brutal honesty? No. That raises the stakes of the reply, so small doubts get filtered out, and small doubts are the useful part. “Brutal” also invites performance instead of accuracy. Ask for one specific piece of work: the strongest case against, or who would object and what they would say.

How do I ask for advice without sounding like I am fishing for approval? Lead with what is undecided, in one sentence, then stop talking. Fishing sounds like a finished story with a question mark stuck on the end; a real request sounds like an unfinished one. If you have already booked the flight, say so and ask what to watch for instead of asking whether it was smart.

Most of the fix sits on your side of the table, which is the part nobody enjoys. People answer the question you built, and a question built around a decision you have already made can only return approval. Take the leaning out. Say what is still open, ask for the mechanism of failure rather than the verdict, and the same friend who has agreed with you for six years starts producing objections worth having. They were there the whole time. Nobody had asked for them in a form that was worth the trouble. If you want to hear the strongest case against a plan before you take it to people who have a stake in your answer, you can rehearse it with Lona, an AI you talk to rather than a friend you might be leaning on. It will not replace the person whose honest objection you need, but it holds the thread from one conversation to the next while you work out what is still open.