Look at any project you are responsible for and count the people whose help you need but cannot order: the engineer in another team, the budget holder two floors up, the peer whose priorities collide with yours, your own manager. The list is usually most of the list. Authority is the small slice of cooperation you can demand; influence is how you get the rest.

The quick version

  • Influence without authority is getting cooperation from people who don't report to you, peers, other teams, senior stakeholders, your own boss, without being able to instruct them.
  • The most-taught model, from Allan Cohen and David Bradford, treats it as exchange: people help you when you help them get something they value, and the unspoken accounting of those favours is the law of reciprocity.
  • What people value isn't only money or favours. Cohen and Bradford call these "currencies", and they include things like recognition, useful information, a sense of purpose, and being treated as an ally.
  • The trap is treating it as manipulation. Done well it is the opposite: you start by understanding what the other person is actually trying to achieve, then find an honest overlap with what you need.

The idea in depth

The cleanest theory of influence without authority comes from Allan R. Cohen (Babson College) and David L. Bradford (Stanford). Their argument, first set out in the Organizational Dynamics article "Influence Without Authority: The Use of Alliances, Reciprocity, and Exchange to Accomplish Work" (1989) and expanded in their book Influence Without Authority, rests on a single mechanism: exchange. An organisation is a web of interdependent people, each with their own goals and pressures. You get cooperation by trading something the other person wants for the help you need. Underneath it all sits what social scientists call the law of reciprocity, the deep human tendency to repay what we receive, good or bad, over time.

So the first move is not the ask. It is the diagnosis. Before you request anything, work out what the other person is measured on, what pressures they're under, and what they're trying to get done. The favour you want is your problem, not theirs, and the request only lands when it also solves something of theirs. That reframing, from "how do I get them to do this?" to "what do they need that I can help with?", is most of the skill.

The currency of exchange is rarely cash. In their chapter "Goods and Services: The Currencies of Exchange", Cohen and Bradford group what people value into five families: inspiration-related (vision, a sense their work matters), task-related (resources, information, a hand with the actual work), position-related (recognition, visibility, reputation), relationship-related (acceptance, inclusion, understanding), and personal (gratitude, autonomy, feeling effective). The practical point: you almost always have currencies to offer that cost you little and matter to them a lot. A public thank-you in front of someone's boss, an early heads-up on a decision, being looped into the interesting work, these buy more goodwill than they cost.

flowchart LR
  A(["Inspiration
vision, meaning"]) --> M(["What does
this person value?"])
  B(["Task
resources, info, help"]) --> M
  C(["Position
recognition, visibility"]) --> M
  D(["Relationship
inclusion, understanding"]) --> M
  E(["Personal
gratitude, autonomy"]) --> M
  M --> O(["Offer a currency
that costs you little,
matters to them"])
Cohen & Bradford's five families of "currency", the things people trade for cooperation. Leaders Loop

Cohen and Bradford turn this into a sequence you can actually run, summarised well by MindTools: assume the other person is a potential ally rather than an obstacle; get clear on your own goals (and set your ego aside); diagnose their world; identify the currencies that matter to each of you; tend to the relationship before you trade on it; then make the exchange. The order matters. People who jump straight to step six, the ask, and skip the diagnosis are the ones who complain that influence "doesn't work on my organisation."

Authority is the cooperation you can demand. Influence is how you get the rest, and the rest is most of it.

An honest limitation. The exchange model is powerful but it is a frame, not a law of physics. It can curdle into transactional book-keeping, tallying favours, expecting an instant return, which corrodes the very relationships it depends on. Reciprocity also works on a slow, fuzzy clock: genuine influence comes from a reservoir of goodwill built long before you need it, not a tit-for-tat ledger settled the same week. And the model says less about power than it pretends: when someone can simply veto you and feels no need to trade, no amount of currency-matching substitutes for going up a level or changing the game. Treat exchange as your default operating system, not your only one.

Where persuasion fits, and where it doesn't

Exchange explains why someone helps you; persuasion shapes how you ask. The two are complementary, and it helps to keep them separate in your head. Robert Cialdini's well-known principles of persuasion, reciprocity, commitment, social proof, liking, authority, scarcity, are largely about the psychology of a single request (we cover them in persuasion principles). They are sharp tools, and the reciprocity principle overlaps directly with Cohen and Bradford's exchange model. But they can be used cynically, which is exactly why the influence-without-authority literature insists on the diagnosis first: technique aimed at someone whose interests you haven't understood reads as manipulation, and people remember being manipulated.

So the move is to lead with substance and let technique serve it. Frame your request in terms of the listener's goals, not yours; bring the evidence that an analytical stakeholder needs and the story that a relational one responds to; and make the small, reasonable ask before the large one. The honest version of every persuasion principle is the same principle used on someone whose problem you are also solving.

A worked example

Take a product manager, call her Priya, who needs three weeks of a platform engineer's time to ship a feature her customers keep asking for. She has no authority over the platform team; their backlog is owned by a different director, and her polite Slack request has been "prioritised" into oblivion for a month. (Illustrative example; the people and figures are made up to show the model, not real cases.)

The instinct is to escalate, to get her director to lean on theirs. Sometimes that's right. But Priya runs the diagnosis first. What is the platform lead actually measured on? Reliability and reducing the flood of one-off integration requests that interrupt his team. What does he value? Fewer interruptions, and visible proof that platform work drives revenue, a story his own leadership rarely hears. That is the overlap. Priya reframes the ask: her feature, built on the platform team's new shared service, becomes the flagship example that justifies the very service he's trying to get adopted. She offers task currency (her team documents the integration so the next five teams self-serve) and position currency (a joint write-up to both leadership groups crediting platform's work).

flowchart TD
  A(["The ask:
3 weeks of an engineer
I can't command"]) --> B{"Diagnose: what is
the platform lead
measured on?"}
  B -->|"Fewer one-off
requests; proof of impact"| C(["Find the overlap:
my feature = his
flagship adoption story"])
  C --> D(["Offer currencies:
docs that cut future
requests + shared credit"])
  D --> E(["Cooperation earned,
not escalated,
and a reusable ally"])
The reframe that earns the help: turn your problem into a solution for theirs. Leaders Loop

The engineer time appears, not because Priya found the right words, but because she found the genuine overlap and made the exchange easy to say yes to. The bonus is durable: she hasn't spent a favour, she's built an ally she can return to. Escalation would have got her the three weeks once and a resentful team forever. Exchange got her the three weeks and a relationship.

Frequently asked questions

Isn't this just a polite name for manipulation?

It's the opposite, when done as intended. Manipulation hides your real aim and exploits the other person's psychology against their interest. The exchange model starts by understanding what they're genuinely trying to achieve and finding an honest overlap, they get something they actually value. The test is simple: would the other person feel used or well-served if they could see your full reasoning? If it's the former, you've crossed into manipulation.

What if I have nothing to offer someone senior?

You almost always have more currency than you think, and the valued currencies are rarely material. Senior people are often starved of honest information from the front line, a clear story they can repeat to their bosses, recognition that reflects well on them, or simply a problem taken off their plate. Diagnose what they're under pressure to deliver, then offer the piece of that you can supply. Managing the relationship upward is a discipline of its own (see managing up, down & across).

How is this different from negotiation?

They're close relatives. Negotiation usually has a defined deal and a table; influence without authority is the everyday, lower-stakes version that happens in corridors and one-to-ones, often with no explicit "deal" named. Both turn on understanding the other side's interests rather than their stated position. If your situation has clear trade-offs and a walk-away point, you've crossed into negotiation proper (BATNA, ZOPA, interest-based bargaining).

Doesn't all this exchange make work feel transactional and cold?

It can, if you run it as a ledger. The healthier mental model is a reservoir of goodwill: you give freely, help before you're asked, and build relationships well before you need anything, so that when you do ask, reciprocity is already in your favour and nobody is counting. The people with the most influence rarely seem to be "doing influence." They've simply banked years of being genuinely useful to others.

When should I escalate to authority instead?

When the cost of building influence outweighs the stakes, when someone is acting in bad faith, or when a genuine priority conflict can only be resolved by whoever owns both priorities. Escalation isn't failure, using a sledgehammer for every nut is. Reserve formal authority for when exchange has honestly been tried, or when the decision genuinely sits above your pay grade and pretending otherwise just wastes time.

Related in the Toolkit

Influence without authority is the everyday cousin of the more formal skills in this area, it shares its engine with persuasion principles, and when the stakes and trade-offs get explicit it becomes negotiation.

Where to go next