Where I’d start

Pick one recent commitment that slipped and work through it with the people involved. Find out what changed after the date was agreed, and how early anyone could have known. Then agree what you’ll do differently for the next commitment.

  • Make it safe to tell you that a date is becoming unlikely.
  • Keep a clear distinction between work you’re exploring and commitments you’ve made to customers.

Talk through one missed commitment

As CEO, you may be hearing different accounts from sales, product and development. Everyone has a reason their part of the work felt urgent. Bring the people involved together and choose one recent example that they remember well.

Ask what you believed you were promising at the time. Was the team estimating a well-understood change, or had a date been attached to an idea that still needed investigation? A date can travel from an internal planning conversation to a customer promise without anyone explicitly deciding to make that commitment.

Write down what happened after that: changes in scope, interruptions, discoveries and time spent waiting. Give people room to explain their decisions. If the conversation becomes a search for someone to blame, you’ll make the next warning harder to raise.

Look at the decisions you control

A team can be working hard while its plan changes every week. If you regularly ask for a customer request to jump the queue, make the trade-off visible. Ask what moves back before accepting the new work.

It helps to keep one shared view of the current priorities. When sales, product and the founder each give developers a separate list, the team has to make business decisions quietly just to get through the day.

You may need to leave room for support and maintenance. Look at how much time that work has actually taken recently. A plan that assumes every available day goes into new features is likely to disappoint you again.

Ask what makes the date uncertain

Some work becomes clearer only after someone starts investigating it. Ask the team which assumptions they’re relying on and what they need to learn next. Give that investigation a small, explicit allocation of time.

For example, a customer integration might depend on an external system behaving as its documentation suggests. A short test with realistic data can tell you more than another meeting about the final delivery date. This is an illustrative example; the useful test will depend on your work.

Where a range is more honest than a single date, use a range. Explain what could move it. For a genuinely fixed deadline, agree which parts of the scope can change and who can make that call.

Make the regular conversation useful

Ask what has reached users since you last spoke and what is still waiting. For the next customer commitment, ask what could put it at risk and whether anyone needs a decision from you.

Agree when the team should raise a concern. You want to hear while there are still choices available. Responding calmly to an early warning is one of the most useful things you can do to make the next warning arrive sooner.

Keep the conversation short enough to repeat. A shared note about the commitment, its uncertainties and any decision needed is often enough. The detail should help someone act.

Try this over the next month

Choose one upcoming piece of work. Agree what “ready for the customer” means and who will confirm it. Record the assumptions behind its timing, then check in regularly as the work moves.

At the end, compare what happened with what you expected. Did the team learn about a risk earlier? Did someone make a needed decision sooner? Did the customer hear about a change with enough notice? Use that conversation to decide what to keep doing.

Repeated surprises after you’ve addressed priorities and interruptions may call for a closer look at technical problems, skills or leadership. Bring in help with a clear question and involve the team in answering it.