When to say no to a customer commitment

A customer wants a feature by a specific date. Sales has commercial incentive to say yes. The engineering leader has to decide whether saying yes is the right call — and if it isn't, how to say no in a way that keeps the relationship.

Written May 2025 from experience running engineering teams with commercial commitments.

When yes is the wrong answer

• The date requires cutting corners you'll regret. The feature will exist by the date; the operational cost of shipping it that fast will exceed the revenue from the commitment.
• The commitment blocks another commitment. Every yes to a specific customer is a no to some other work, and sometimes the other work is more valuable.
• The team's morale is already stretched. Another crunch project degrades everything else the team is trying to do.
• The customer is willing to accept a compromise. They're asking for the maximum; they'd be happy with less.

How to say no well

• Say it early. The customer will accept a considered no in month one better than a scrambled explanation in month three.
• Propose an alternative. "We can't do X by that date. We can do Y by that date, or X by a later date." Give them a decision to make.
• Be specific about the constraint. "This requires work we haven't scoped, and rushing it produces reliability risk we won't accept" is a legitimate reason and a defensible one.
• Let the commercial side make the call. Once you've laid out the trade-offs, the decision to push back or agree to the alternative is a business decision, not just an engineering one.

What saying no does

• Preserves the team's ability to ship reliable software.
• Builds trust with customers who eventually see that your yeses actually ship on time.
• Creates space for the engineering work that isn't customer-driven but is necessary.

Saying yes feels helpful. Saying yes when you shouldn't is a form of harm.