PM DEPOT

Back to Stakeholder Management
BUREAU
OF S·H
Bureau of Stakeholder Hazards  ·  A Field Dispatch Dispatch N° III
Hazard Class · Unauthorised Commitment

Someone Committed Your Team to a Deadline You Didn't Agree To

Sales closed the deal by promising the feature. Marketing announced the date. Your CEO told a partner. You were not in any of these rooms. A guide to recovery without arson.

· · · · ·

You are eating a sandwich at your desk. A customer replies to a thread with cheerful gratitude — “we're so excited about the March 15th launch of X, our whole team is ready.” March 15th. Of X. Neither of which you have ever agreed to, neither of which engineering has scoped, neither of which exists in any document you have ever read. You put down the sandwich. The sandwich is over.

Welcome to the third great hazard: the unauthorised commitment. Somewhere, a human being who is not you promised one of your team's resources — time, scope, or a specific feature — to another human being who is now operating in good faith on that promise. You are now in a situation where the commitment exists, the expectation exists, and the engineering capacity does not.

The instinct, which you should resist, is to march into a meeting with a list of grievances. The correct reaction is to slow down for forty-five minutes and figure out what actually happened, because unauthorised commitments come in at least three flavors, and each needs a different response.

The Three Species

The Sales Commitment

An account executive, under pressure, closed a deal by promising a capability your team didn't know about. This is the most common species. The committer was acting in what they believe is the company's interest and was rewarded for doing it. They are not malicious; they are incentivised.

The Executive Promise

A senior leader — often in a board meeting, investor call, or partner dinner — named a date or feature aloud. This is the most dangerous species because the promise now has social weight attached to a senior person's credibility, and unwinding it means asking that person to walk something back publicly.

The Peer Overreach

A counterpart in marketing, design, or another product team made a decision that assumes your team's involvement without checking. Often done with the phrase “I assumed we'd just…”. This is the most irritating species, because the committer has no authority to commit you and also doesn't realise that's a problem.

The First Forty-Five Minutes

Before you do anything else — absolutely anything — answer three questions, in writing, for yourself.

  1. What exactly was promised? The literal words. Not your summary, not the customer's summary. Get the original email or deck. Half the time the commitment is vaguer than it felt in your stomach, and you are panicking about your own interpretation.
  2. Who committed it, and did they have authority? “Sales told a customer” is not the same problem as “the CRO told a customer.” Authority matters because it determines who has to fix it.
  3. What is the real cost of delivery? Not the scared cost. The actual cost, after talking to your tech lead for eight minutes. Sometimes the commitment is ugly but cheap. Sometimes it is impossible. These paths diverge at this fork, and you cannot pick the right response until you know which one you're on.
Contrarian note A surprising number of unauthorised commitments are actually gifts. They surface constraints that your team had been dodging — dates you were avoiding setting, features you had been quietly deprioritising, conversations you hadn't wanted to have. The commitment is now a forcing function. Hate the process, use the forcing function.

The Fix — The Three-Option Move

Do not go to the committer and say “you can't promise that.” This is correct but useless. Instead, go to the committer with three paths, costed, and make them pick.

  1. Option A — Deliver the commitment as stated. Here is what we would have to stop doing to make it work. These teams absorb the impact. This feature slips by N weeks. Here is the risk.
  2. Option B — Partially deliver. The part of the promise we can do by that date is X. The part that slips is Y. Here is how we'd frame that to the customer/exec/partner so it still feels like a win.
  3. Option C — Renegotiate the date. The committer goes back and extends by N weeks. Here is a script that gives them cover.

The magic of this move is that the committer now has to choose a cost, and choosing a cost is what committing should have done in the first place. You have not said no. You have converted the unauthorised commitment into an authorised one, by escalating the question to the person who actually has the authority to make it.

Phrases That Work

I just saw the commitment to the customer — want to get aligned on how we land this without over-promising. I've sketched three options, with costs. Can I grab twenty minutes?
To make the March 15th date work, we'd pause the onboarding project and ship a smaller version of what was promised. Want to walk through the trade-off before I take it to engineering?
I don't want to put you in a bad position with the customer. Here's the language I'd use to reset expectations gracefully — does that feel safe enough?

The Long Game

Unauthorised commitments happen because your team's availability, priorities, and constraints are invisible to the rest of the company. The long-term fix — the actual fix — is not a better process. It is radical visibility.

Publish, monthly, a short “what we are and aren't doing this quarter” doc to every adjacent team: sales, CS, marketing, leadership. One page. Two lists. The goal is that the next time someone is about to commit your team to something, the mental friction of “wait, are they doing that right now?” is slightly higher than the friction of “I'll just ask Jess first.”

You cannot stop unauthorised commitments by being annoyed after the fact. You can make them rare by making your team's reality legible before the fact. Legibility is the quietest superpower in stakeholder management.

Working rule: promises without trade-offs are not promises, they are hopes. Your job, in the forty-five minutes after you read that customer email, is to put a trade-off next to the hope. That's recovery.

← Back to Stakeholder Management