Influence without authority: what you can trade, and when to stop asking

You are asking someone to do work that their manager did not ask them to do. That is the whole problem, stated exactly. Everything else about this subject is a consequence of it.

Most advice on the topic answers a different question: how to be liked, how to build rapport, how to tell a compelling story. Those are not nothing, but they are not the binding constraint.

The person you are asking is not refusing because they dislike you. They refuse, usually silently and by deferral, because your request sits behind six things someone who writes their review already put in front of it.

What to take away

Why the request keeps getting agreed to and not done

There is a specific failure that anyone working across teams will recognize. You ask, they agree, they mean it, and nothing happens. Then it happens again the following month.

That is not dishonesty and it is rarely even reluctance. Agreement in a meeting costs nothing and is the cheapest way to end a conversation politely. Doing the work costs a slot in a week that is already full. The agreement was real; it just never had to survive contact with a prioritization call it was never entered into.

So the useful question is never "did they say yes". It is "whose list did this get onto, in what position, and who else can see that list". If the answer is that it lives only in your notes and in their intention, you have not got a commitment. You have got a pleasant conversation.

Find out what they are measured on, and believe it

The single highest-return piece of work here is unglamorous: find out what the person and their team are actually judged on, and then check whether your request helps or hurts that.

Not what their function is nominally for. What gets counted, what gets reported upward, and what would make their quarter go badly. Those are often different from the official purpose, and where they differ, the counted thing wins.

Once you know it, one of three things is true, and each has a different move.

Your request helps what they are measured on. Then you are not asking a favor, you are handing them something, and the work is simply to make that visible rather than to persuade. Say the connection out loud in the first two sentences. People are surprisingly bad at spotting that an outside request serves their own numbers.

Your request is neutral. Then you are asking for a favor and should say so plainly. Favors are a real and legitimate currency, they are finite, and they are spent badly by people who dress them up as something else.

Your request costs them. Then no amount of framing will fix it, and pretending otherwise damages you when they work it out. This is the case where you either change the request, find whoever is willing to trade the cost against something, or drop it. Recognizing this case early is worth more than any technique.

The three things you actually have to offer

Without authority you cannot instruct, so you are trading. The inventory is short.

Information. You usually know something they do not: what is coming, what another team is planning, why a decision went the way it did, what a customer actually said. The gap between what each side knows is the one asset that does not depend on your standing. It is genuinely valuable and most people give it away too cheaply or, worse, sit on it as though hoarding it were an advantage. It is not an advantage. Being the person whose information turns out to be accurate is the advantage.

Absorbed work. You can make the request smaller. Do the preparation, write the first draft, gather the inputs, book the room, take the follow-up. Every unit of work you take out of their week is a unit of resistance removed, and it is the only offer here that scales with effort rather than with standing.

Cover. You can make it safe to say yes. Put the reasoning in writing so they are not the only person exposed if it goes wrong. Get the sponsor to say it in the meeting rather than reporting that the sponsor said it. Take the blame publicly if the request turns out to have been a bad idea. Cover is the currency people are least willing to spend and it moves things nothing else moves.

Notice what is absent from that list: enthusiasm, urgency, and the strength of your argument. Those affect whether you get a hearing. They do not affect the arithmetic once you have one.

Make the ask small enough to survive a bad week

A large request is refused by a busy person almost automatically, because it cannot be assessed quickly and the safe response to an unassessable request is to defer it.

Do not shrink the goal, but shrink the first step until it is obviously answerable.

Ask for a decision rather than a project, and for thirty minutes and a name rather than a commitment. Ask them to tell you why it will not work; most people accept that easily, and it gives you the objection or the agreement.

The other half of this is a date and an owner. "We should look at that" is where these things go to die. "Can you tell me by Thursday whether your team can take this in October, and if not who should I ask" is a question with a shape, and shaped questions get answered.

Write it so it survives being forwarded

You will almost never be in the room where your request is actually decided. Someone else will represent it, from memory, in one sentence, badly.

That is the design constraint on how you communicate here, and it argues for something short and written. The federal guidance on writing for understanding is written for public notices, and the same discipline applies: short sections, active voice, the decision stated rather than implied.

If those four fit in a paragraph someone can paste into a message, a person who has not spoken to you and does not care much can argue for your request.

Anything that depends on your delivery, your tone, or your presence in the room will not travel. Assume it will be repeated by the least invested person who heard it.

The account you are drawing on

There is a real thing behind all of this that resembles credit, and it is built from a boringly small number of behaviors.

Being right about things that were checkable later. Saying you do not know when you do not know, which makes your other statements worth anything. Not asking for what you do not need.

Not escalating around someone without telling them first. Delivering the part you promised, especially the unglamorous part after the interesting part is done.

The account is depleted by asking often, by asking for things that turn out not to have mattered, and by being unreliable on small things, which people generalize from far more than they generalize from large ones.

The main practical consequence is that most of the work happens before the request. The person deciding whether to spend a week of their team's time on your behalf is answering a question about you, and they answer it from a history you cannot construct on the day.

Knowing when to stop

Persistence past the point of usefulness is the most common way people damage themselves here, and it is done in good faith by people who have been told that persistence is a virtue.

Three signals are worth taking seriously. The request was agreed to twice and not done twice, so the agreement is not where the decision lives. The objection changes each time you address it, which usually means the stated objection was never the real one.

The cost genuinely falls on them, the benefit genuinely falls elsewhere, and no one with authority over both is willing to say the trade is worth making.

In all three, the move is the same. Stop asking the person and start describing the trade to whoever owns both sides of it.

If you tell them you are doing this and why, you are not going over their head. It is the only honest way to resolve what a person without authority cannot, and it beats six months of pleasant meetings.

Common questions

Is this just office politics?

It is the part of coordination that no organization chart handles. Every structure leaves work that crosses lines nobody owns, and someone has to move it. Politics is when the mechanism is hidden. Saying out loud what a request costs, who benefits, and who has to choose is the opposite of that.

What if I genuinely have nothing to offer?

You almost always have absorbed work, which requires no standing at all. If the request is small enough and you have done the preparation, the remaining cost may be an hour and a signature. Start there and build the account from things that actually got delivered.

How do I handle someone who always says yes and never delivers?

Stop treating their yes as information. Move to specifics they cannot agree to vaguely: a date, a named person, a next step you can observe. If those are also agreed and missed, you have your answer, and the conversation to have is with whoever prioritizes their work, conducted openly rather than as a complaint.

Should I get my manager to ask instead?

Sometimes, and it works, and it costs. Borrowed authority gets the thing done once and teaches everyone that your requests arrive with an escalation attached. Keep it for the cases where the trade is genuinely a management decision, which is exactly where it belongs.

Does any of this change if I later get authority?

The mechanics do not, but the failure mode inverts completely. Without authority your problem is being ignored. With it, your problem is that everything you say gets over-weighted and the information coming back to you quietly degrades.