Influence Without Authority

Almost nobody who blocks your work does it because they were not convinced.

Think about the last time another team did not do the thing you needed. Run through the likely reasons: they had four other priorities, the request arrived without a deadline they were accountable to, doing it would have made one of their own numbers worse, or the person who could say yes never actually saw the request.

Notice what is absent from that list. "They heard your case and found it unpersuasive" is rarely the explanation. Which means a better case, more forcefully argued, addresses a problem that was not there.

The structural problem underneath

Functions are deliberately given competing objectives. Sales is measured on speed, risk on certainty. Engineering is measured on stability, growth on shipping. That tension is not dysfunction — it is the design. Which means a cross-team disagreement is usually two people each correctly optimising for what they are paid to optimise for.

You are not going to resolve that by being right. You resolve it by finding the version of your request that does not force the other person to lose.

This is the entire skill. Everything else is execution on top of it.

Most people know roughly what a neighbouring team does. Very few know what that team's manager is asked about in their own performance review. That second thing is the one that determines what gets prioritised, and it is usually one question away.

Before I ask you for anything — what does a good quarter
look like for your team?

And what's the thing that would make it a bad one?

People answer this readily, because it is a question about them and almost nobody asks it. Two minutes of listening here is worth more than a month of follow-up emails.

What to listen for

• The number. Uptime, cycle time, pipeline, defect rate, cost per unit, ticket backlog. There is always one.
• The constraint. What they are short of — usually people, sometimes information, occasionally a decision nobody will make.
• The recurring pain. The thing they complain about every time it comes up. This is your opening.
• The thing they get blamed for. Not the same as their metric, and often more powerful. People will move fast to avoid a specific known blame.

Same work. Different framing. Wildly different response rate.

Once you know their number, the request rewrites itself. The move is to stop describing what you need and start describing what it does for them — and where it genuinely does nothing for them, to say so honestly and pay for it another way.

That last row matters more than the other three. Naming honestly that something does not benefit the other team buys an unreasonable amount of credibility, because it establishes that when you do say something helps them, you are not making it up.

The three currencies you actually have

1. Information. You know something they cannot see — customer behaviour, an upstream change, a decision forming elsewhere. Being someone's early-warning system is a real asset, and costs you almost nothing.
2. Access. You can get them in front of someone, into a forum, or onto a distribution list they are not on.
3. Relief. You can take work away. The strongest opening move available is arriving with something you have already done for them.

Two axes, four groups, four different moves. Most wasted effort comes from treating everyone the same.

Take the people who touch your initiative and place each on two axes: how much power they have over whether it happens, and how much they care about the outcome. The four quadrants call for genuinely different behaviour, and confusing them is expensive.

Find the actual decider

The person with the relevant title is frequently not the person whose opinion settles it. Ask plainly, before you invest weeks: "If this were going to get approved, who would need to be comfortable with it?" People generally answer honestly, and the answer is often someone you had not thought about.

Influence accrues on a schedule. Knowing the schedule stops you concluding it is not working.

Practitioners who teach this consistently describe the same rough arc: roughly three to six months to establish baseline credibility with a team you have no authority over, and something like twelve to eighteen months before you have durable influence — the kind where people bring you in before decisions rather than after.

Accelerate with early, visible wins

• Pick a first collaboration that is small and almost certain to succeed. The point is not the outcome, it is establishing the pattern that working with you goes well.
• Deliver whatever you promised, on the date you promised it, including the boring parts. Reliability compounds faster than brilliance in cross-team work.
• Give the credit away publicly. It costs you nothing measurable and it buys the next three requests.
• When something slips, say so before they notice. One early warning is worth more than three successful deliveries.

One useful standard for cross-functional work: aim for something like eighty percent agreement and full commitment, rather than complete alignment. Pushing for the last twenty percent of consensus is where most initiatives stall, and it is usually not necessary to move.

In matrixed organisations, the person with the clearest written record usually wins the ambiguity.

Cross-team commitments made verbally in a meeting have a short half-life. Everyone leaves with a slightly different version, priorities shift the following week, and by the time it matters nobody can reconstruct what was agreed. The person who writes it down decides what was agreed — not by manipulation, but simply because theirs is the only record.

Subject: [Project] — what we agreed today

Quick capture so we're all working from the same thing.

Agreed:
  - [Team A] will [specific thing] by [date]
  - [Team B] will [specific thing] by [date]

Still open:
  - [the thing nobody committed to — name it]

Assumption I'm working on: [the thing you inferred but nobody
said out loud]. Tell me if that's wrong.

[name]

The two structural elements are the "still open" line and the stated assumption. The first prevents an unowned item from quietly becoming yours. The second surfaces disagreements that would otherwise appear six weeks later as a surprise.

Informal influence has limits. Recognising them late is worse than escalating early.

There is a point at which continuing to work the relationship is just absorbing the cost of a decision nobody will make. Harvard's guidance on this is blunt and correct: when repeated informal effort stalls, get visible support from a sponsor, and in persistent cases get the authority formalised.

Signals it is time

• You have asked three times across at least two months and the answer has been "next quarter" each time.
• The blocker is a resourcing decision that the person you are talking to does not have the authority to make.
• The other team's incentives are genuinely opposed to yours, and no reframing changes that. This is a leadership decision, not a relationship problem.
• You are absorbing the delay silently, and your own commitments are slipping because of it.

How to escalate without burning the relationship

1. Tell them you are going to. This single step is what separates escalation from going behind someone's back. Almost always survivable when done in advance, almost never when discovered afterwards.
2. Escalate the decision, not the person. Frame it as two priorities that need someone senior to sequence, not as a team failing to cooperate.
3. Bring the trade-off, not the complaint. Senior people respond to "here are the two options and what each costs." They disengage from "we are blocked."
4. Do it once. Repeated escalation converts you into the person who cannot resolve things at their own level.

I don't think either of us can solve this at our level —
your priorities and mine are genuinely pointing different
ways and neither of us is wrong.

I'm going to raise it with [name] and ask them to sequence
it. I wanted you to hear that from me rather than find out.

Anything you want me to represent accurately about your side?

That final question is what makes it collaborative. You are not reporting them; you are asking them to help you describe their position fairly. People rarely object to that.