Pushing back on an unrealistic deadline: three offers, not one no

The date usually arrives before the work is understood. A launch, a board meeting, a customer who was told "end of the quarter". You look at it and know it will not hold, and the word that comes to mind is "unrealistic".

That word is where most pushback fails. "Unrealistic" is your conclusion. The person who set the date has a fact (the date) and hears an opinion against it. Pushing back works when you arrive with two things they do not have: a figure someone owns, and a choice they can make.

First find out what kind of date it is

Three different things are called a deadline, and they bend differently.

• An external date. A regulation takes effect, a contract has a penalty clause, an event happens on a day. The date is real and will not move. Scope has to.
• A promised date. Someone inside has already told a customer or a board. The date can move, at a cost somebody has to carry in person. That is a commitment made before the evidence, and it should be recorded as one.
• A preferred date. It was chosen because round dates create urgency. Nothing happens the day after. This one moves the moment anyone asks.

One question sorts them: "What happens on the day after, if it is not done?" Ask it plainly and write down the answer. A surprising share of hard deadlines turn out to be the third kind once they have to be put in a sentence someone else will read.

Answer with a figure that has an owner

Do not push back in the meeting where the date appears. You do not have a number yet, and a manager's instinct against an executive's date loses. Say when you will have one: "I can tell you on Thursday what fits by then." Then get the size from the team that would do the work, with how sure they are. A figure with a confidence level attached is evidence. A figure you guessed on the spot is a second opinion, and it will be quoted back to you as a promise.

Two days is usually enough. If the team honestly cannot size it in two days, that is the finding: the date was set for work nobody understands yet, and the first thing to agree is a short, time-boxed look, not a delivery date.

Bring three offers, not one no

"We can't" ends the conversation with the date still standing. Offers keep it going and move the decision to the person who owns the trade. There are only three.

1. The date holds and the scope shrinks. Name what ships by then and, more important, what does not. "The export works for the two largest customers; the self-serve settings page comes after."
2. The scope holds and the date moves. Name the date the team's figure supports, and what it depends on.
3. Both hold and something else gives. Name the commitment that would be displaced, and who was promised it. This is the offer people forget, and often the one that gets chosen, because the person asking usually outranks the thing in the way. It is still a visible trade, not a squeeze.

Adding people is a fourth offer only early in the work and only when the work divides cleanly. Late, it makes the date worse. Say that too, once, so it does not come back as a suggestion in week six.

What to say

I've taken the 30 June date to the team. Their figure for the full scope is four FTE-months, medium confidence, and we have two and a half before then. Three ways this can work:

1. Keep 30 June, ship the export for the two largest customers only. Settings page follows in July.

2. Keep the full scope, land on 15 August.

3. Keep both, and the billing migration moves to next quarter. Finance was promised that, so I'd tell them this week.

Which of these do you want? If none of them works, tell me what I'm missing about the date.

There is no adjective in it. Nobody is called unrealistic. The last line matters: sometimes there is a fact you did not have, and the date really is the first kind.

When the answer is still "the date stands, all of it"

It happens, and sometimes it is the right call: the deal is worth the risk. What you can still insist on is that the risk is taken in the open. Commit, and record it as an early commitment with four things written down: why now, what is missing, who accepts the risk by name, and what would change the decision.

This is not a way of saying "I told you so" later. It does two practical jobs. The person who accepts the risk reads the sentence with their name in it, and sometimes the date moves at that moment. And at the end of the quarter the miss is counted as a decision that was made knowingly, instead of as a delivery failure by the team that warned about it.

What not to do

• Agree and hope. The date does not get more likely, and the warning you did not give is now yours.
• Say "we'll try". It is heard as yes by the person asking and as no by the team.
• Re-estimate downward. The work did not shrink because the plan needed it to.
• Pad everything else to make room. Hidden slack is found, and after that none of your figures are believed.

And if you accepted a date that is now slipping, the same rule applies in the other direction: say it early and with a new date, not in the last week.

If you use DeliverySheet

The product holds the two records this conversation needs, and leaves the conversation to you. On the standard path a demand cannot be committed until the delivering team has given a capacity figure, with a confidence level per row, so the figure in your message has an owner. Committing before that is allowed and requires a commitment exception: why now, what is missing, who accepts the risk and what would change the decision, all four required. Those commitments are listed in their own report, Early commitments, with the outcome each one ended on. If the date or the scope changes later, that is a recorded re-baseline with its reason, not an edit.