A Deadline Has Two Costs
A deadline conversation starts as if the team is trying to discover the correct date. Engineering estimates the work. Sales brings the customer pressure. Then someone asks both sides to be realistic.
The problem is that each side is being realistic about a different failure.
Engineering pays when the date is too early. The team absorbs rushed decisions, missing edge cases, rework, and the reputational cost of shipping something unreliable. More time reduces uncertainty and gives the team room to build the version it can defend.
Sales pays when the date is too late. The company loses momentum, delays revenue, and makes customers wait while their problem remains unsolved. More time may improve the product, but it can also make the product commercially irrelevant.
My rule for deadlines is that everyone should be unhappy. Engineering should think the date is too soon. Sales should think it is too late.
Comfortable Dates Are Departmental Dates
If Engineering is comfortable with the deadline, the plan may be treating market time as free. The date protects technical certainty while asking the rest of the company to absorb every delay.
If Sales is comfortable, the plan may be treating implementation risk as free. The date protects the deal while asking Engineering, Support, and eventually the customer to absorb the consequences.
Neither date is a company deadline. It is one department's preferred outcome presented as a shared plan.
A company deadline has to price both costs. That does not mean everyone leaves the room equally satisfied. It means neither side gets to pretend its risk is the only real one.
Unhappiness Is a Signal, Not a Method
This rule does not mean taking the Engineering date, taking the Sales date, and choosing the midpoint. Dates are not made credible by averaging two incompatible assumptions.
The useful part of the disagreement is the information behind it. Engineering needs to explain which uncertainty remains, which work is required, and what fails if the date moves forward. Sales needs to explain what changes commercially, which commitment matters, and what is actually lost if the date moves back.
Once both sides make the costs concrete, some urgency disappears. A vague customer expectation is not the same as a signed commitment. A desire for more polish is not the same as a reliability requirement. The deadline improves when both sides stop hiding preferences inside estimates.
The remaining discomfort is valuable. It means the decision still reflects both the cost of waiting and the cost of rushing.
Change Scope Before Believing a Bad Date
Sometimes Engineering's earliest credible date comes after Sales' latest useful date. No scheduling technique can repair that gap. The scope is wrong.
The team can cut work, split the release, narrow the customer promise, or deliver the smallest version that solves the commercial problem. It cannot keep the original scope, keep the original date, and call the result alignment.
Sometimes even the smallest honest version arrives too late. Then there is no good deadline. The company has to reject or renegotiate the commitment instead of forcing a compromise into the calendar.
When a smaller version is viable, a good deadline makes both sides unhappy for different reasons. Engineering ships less than it wanted to build. Sales waits longer than it wanted to wait. The company avoids making a promise that only works if nothing goes wrong.
Scope is the variable that turns disagreement into a plan. Pressure alone only turns disagreement into hidden risk.
The Commitment Has to Survive the Meeting
A deadline matters only if the tradeoff remains stable after the meeting. Engineering commits to the reduced scope and the known risks. Sales commits to selling that scope instead of quietly restoring every removed promise.
Without that discipline, the same negotiation restarts every week. Sales pulls the date forward through customer escalation. Engineering pushes it back as omitted work returns. The deadline becomes a recurring argument instead of a decision.
A good deadline is not the date everyone agrees is reasonable. It is the date both sides are willing to own, even though neither side got what it wanted.