Influence Without Authority

Status Updates That Actually Inform was about keeping the people around you informed so they trust your read on reality. This page is about what you do next: getting those same people — and people you’ve never met — to actually do something for you, when none of them work for you and none of them have to.

Here is the uncomfortable truth about almost any real job: the work that matters most crosses lines you don’t control. The security team, the finance approver, the other department, the vendor, the busy colleague two floors up — you need them, and you can’t tell them what to do. This page teaches you how to get things done anyway, not through cleverness or politics, but through a small number of durable principles: credibility, reciprocity, and shared goals. Master these and you stop being someone who can only move the few things inside your own fence.

Why authority stops at the boundary

Inside your own team, authority does a lot of quiet work. When you ask a direct report to do something, there’s an implicit backstop: it’s their job, you affect their pay and standing, and refusing has consequences. You rarely have to use that backstop — but it’s there, and it makes requests land.

Step across the boundary to someone who doesn’t report to you and that backstop vanishes. You have no lever over their pay, their priorities, or their day. If you ask in the same tone you’d use with your own team — a directive, a “can you just” — one of two things happens. Either they ignore you, or they comply resentfully and slowly, and remember it. Authority you don’t have can’t be borrowed by acting as if you do.

So what does move people across a boundary? Strip it down and it’s three things, all of which you have to build rather than command:

What moves people who don't report to you
------------------------------------------
Credibility   Do they trust that your ask is real, sound, and worth their time?
Reciprocity   Is there a relationship of give-and-take — have you helped them, or will you?
Shared goal   Does helping you also serve something they already care about?

Notice none of these can be produced on demand in the moment you need the favour. They’re a balance you build up over time and spend down when you make an ask. That single reframe — influence is a balance, not a transaction — is the heart of this whole page.

The evidence — the currencies of influence

Studies of how work actually gets done in organizations consistently find that formal authority explains only a fraction of who gets things moved. Allan Cohen and David Bradford’s long-running work on “influence without authority” frames it as an exchange of currencies — help, recognition, information, access, a returned favour — traded across a web of relationships. The practical lesson is durable even if you never learn the theory: people say yes to those they trust and owe, far more than to those who simply outrank the room.

Find the shared goal — or the other person’s interest

The single biggest mistake in cross-boundary requests is framing the ask around your need. “I need this by Friday for my project” tells the other person exactly one thing: that helping you costs them time and gets them nothing. Even if they’re kind, you’ve made yourself easy to deprioritize.

The fix is to do the harder thinking before you ask: what does this person actually care about, and how does helping me touch it? Sometimes there’s a genuine shared goal — you both want the same launch to succeed. Sometimes it’s their own separate interest — they’re measured on system stability, or on closing tickets fast, or on not being blamed later. Either way, the move is the same: frame the ask in their terms, not yours.

The ask, framed two ways
------------------------
Yours:   "Can you review our data change by Friday? We're blocked."
Theirs:  "This change touches the customer records your team owns — I'd
          rather you catch a problem now than get paged about it later.
          Could we get your eyes on it before Friday?"

The second version costs you one extra sentence of thought and changes the whole meeting. You’ve shown you understand what they’re on the hook for, and you’ve connected your need to their world. That is not manipulation — it’s respect. You’re not tricking them into caring; you’re finding the real reason they already might.

In practice — the warehouse and the loading dock

A warehouse manager needed the loading-dock crew — a separate shift, separate boss — to change how they stacked returns so her team could process them faster. Her first ask (“can you stack them this way, it helps us”) went nowhere; it was pure cost to them. So she watched their morning and noticed the dock crew hated the pile-up that built when returns weren’t cleared. She went back: “If we can clear returns twice as fast, that pile by your door disappears by 10am.” Same request, framed around their daily headache. It was done by the end of the week.

Build the relationship before you need it

Here is the timing problem with everything above. Credibility and reciprocity are exactly what you cannot manufacture in the ten minutes before a deadline — and yet that’s precisely when people first try to use them. They show up, hat in hand, to someone they’ve never invested a minute in, and are surprised the well is dry.

The people who are quietly effective across boundaries do the opposite. They spend small amounts of relationship before they need anything: a genuine thank-you, sharing a useful bit of information, showing up to help when they’re the ones being asked, learning the other team’s pressures and names. None of it is calculated in a cynical way — it’s mostly just being a decent, reliable colleague. But it compounds. By the time they have an urgent ask, they’re not a stranger requesting a favour; they’re someone who has a relationship and a track record, making a reasonable request.

Reputation works the same way and travels faster than you think. In most organizations, whether you’re “someone whose asks are worth saying yes to” is a fact that circulates ahead of you. If your past requests were clear, honest, and followed through, people you’ve never worked with will already be inclined to help. If you’ve over-asked, cried wolf, or dropped things you promised, that circulates too.

The common trap — the transactional relationship

The trap is treating relationship-building as a technique you switch on when you want something — being warm to the finance approver the week you need the budget, then going cold. People detect this instantly, and it does more damage than no relationship at all, because now you’ve signalled that your interest is purely extractive. The only version that works is the boring one: be reliably decent to people whether or not you currently need them. Influence is a byproduct of that, not a substitute for it.

Make it easy to say yes

Even someone who trusts you and shares your goal has a limited budget of time and attention. Every request you make lands as a small pile of work: figuring out what you actually want, deciding whether it’s safe, and doing the thing. Your job is to shrink that pile. The easier you make it to say yes, the more often you’ll hear yes.

Three things quietly kill cross-boundary requests, and all three are yours to remove:

• Effort. Don’t send “can we talk about the integration sometime?” Send the specific, small, bounded ask: “Could you spend 15 minutes Thursday confirming these three field names are right?” A request someone can complete in a known amount of time is far easier to accept than an open-ended commitment.
• Ambiguity. If they have to come back and ask what you mean, you’ve added a round-trip and a reason to defer. Say exactly what you want, by when, in what form, and why. Do the clarifying for them.
• Risk. Often the real reason for a no is unspoken: “if I approve this and it breaks, I’m on the hook.” Name and reduce that risk yourself — “I’ve had it reviewed, here’s the rollback plan, and I’ll own it if anything goes wrong.” You’re not just asking for a yes; you’re making the yes safe.

A hard-to-say-yes ask vs. an easy one
-------------------------------------
Hard:   "Hey, can your team help us out with the migration at some point?
         It's kind of urgent."
Easy:   "We need one thing from your team: confirm these 3 records map
         correctly (list attached, ~15 min). Deadline Friday. I've drafted
         it so you're checking, not building. Any risk I've missed?"

The second version has already done the work of thinking. The other person can glance at it, see the shape of the yes, and give it. Every ambiguity you resolve in advance is a ambiguity they don’t have to resolve — and resolving it is often the very thing that gets a request quietly shelved.

Spend credibility carefully — and rebuild it with follow-through

Credibility is a limited balance, and the fastest way to drain it is to over-ask. If every request you make is marked urgent, none of them are. If you escalate small things, ask for more than you need “to be safe,” or pull the emergency lever routinely, people learn to discount you. The next time something genuinely is urgent, you’ll have spent the very credibility you need. Ask for what you actually need, at the priority it actually deserves.

The other half is follow-through, and it’s where cooperation is either earned or destroyed. When someone helps you, two things build your balance for next time: close the loop and give credit. Tell them what their help made possible (“your review caught the bug that would’ve hit us Monday — thank you”). Thank them where their own boss can see it. And most of all, do what you said you’d do with what they gave you. A person who helped you once and saw it land well, and saw you honour your side, will help you again with less hesitation. A person who helped and then watched you drop the ball, or take the credit, has learned not to bother.

Case study — Amazon and the “single-threaded owner”

Large organizations discovered long ago that big goals die in the gaps between teams that don’t report to each other. Amazon’s widely-described answer, developed through the 2010s, was the “single-threaded owner”: one person made clearly responsible for an outcome that spans many teams — with no authority over most of them. The role only works through exactly the skills on this page. The successful owners don’t try to command the teams they depend on; they build the relationships, frame the goal so each team sees its own stake, and make cooperating easy. The lesson generalizes far beyond Amazon: someone almost always owns an outcome larger than their authority, and influence — not the org chart — is what closes the gap.

How it connects

Influence is the foundation the next two pages build on. When shared goals aren’t enough and interests genuinely conflict, you move into Negotiation Basics — which is just influence when there’s something to trade and something to give up. And when influence has honestly been tried and failed, you reach for Escalation Done Right — the deliberate, respectful use of someone else’s authority, which itself depends on the credibility this page is about. It also rests on Mapping Your Stakeholders: you can’t build relationships across boundaries you haven’t noticed exist.

Try this

This week, pick one thing you need from a person or team that doesn’t report to you. Before you ask, spend five minutes on two questions: what does this person actually care about or get measured on? and what would make saying yes cheap and safe for them? Then rewrite your request so it names their interest and shrinks the effort, ambiguity, and risk. Separately, do one relationship deposit with no ask attached — a genuine thank-you, a useful piece of information, an offer to help — with someone you’ll likely need later.

Reflect

1. Think of a recent cross-boundary request that went nowhere. Was it framed around your need or theirs — and what would the “theirs” version have said?
2. Who are two or three people outside your team you regularly depend on? For each, honestly: do you have credibility and reciprocity built up, or would you be asking from empty?
3. Recall a time someone said no (or “later”) to a reasonable ask. How much of that was effort, ambiguity, or unspoken risk you could have removed in advance?
4. Do you over-ask — marking things urgent, asking for more than you need, escalating small things? What has that cost you in how seriously people now take you?
5. When someone last helped you across a boundary, did you close the loop and give credit — or did the favour just quietly disappear? What would they expect if you asked again?

Show reflections

1. Most stalled requests are framed around the asker’s need, which gives the other person nothing to care about. The useful move is to reconstruct their interest — what they’re measured on or worried about — and see how differently the ask would have read.
2. This is a stock-take of your influence balance. Empty accounts aren’t a failure; they’re a signal to start making small deposits now, before the day you need a big withdrawal.
3. A strong answer resists blaming the other person and looks at the pile of work the request created — the open-endedness, the unclear specifics, the unnamed risk. Almost all of it was yours to shrink.
4. Over-asking is invisible to the person doing it and obvious to everyone else. If you can’t remember the last time you didn’t mark something urgent, your credibility is probably already discounted — and the fix is to ration your asks to their true priority.
5. Cooperation is either reinforced or quietly killed by what happens after the help. If favours vanish into silence, people learn helping you is thankless; closing the loop and crediting them is what makes the next yes easy.