What do you do when someone wants one firm date and your honest answer is a range?
answer
- ask what the date is for
- two ends, not one number
- offer a commitment you can keep
- give a date for the decision
- put what you promised in writing
basics
~10 sTests whether you hold a boundary without stonewalling. Ask what the date is for, give a range with the condition that decides it, and commit to something you can keep instead.
how to answer
5 beats- find out what the date is actually forOpen by distinguishing an external commitment from a rough planning question — they deserve completely different answers. One or two sentences; do not turn this into an interrogation of the asker.
- why you will not hand over a single numberSay plainly what you will not do: put one date on a thing you cannot yet size. Then give the two ends in the same breath, so the refusal arrives with something usable attached rather than as a wall. The work in this beat is holding the line without sounding obstructive.
- offer a commitment you can genuinely keepReplace the date you cannot promise with one you can: a scoped slice, a guaranteed decision point, or a checkpoint date. This is the beat that separates holding a line from refusing to engage.
- say when the range narrows and what will narrow itCommit to a specific moment when you will come back with a tighter number, and name what you will have learned by then. Without this, a range reads as a permanent hedge.
- record what you actually promisedClose by putting the commitment in writing where the people planning around it can see it. Verbal ranges tend to be remembered as the optimistic end by everyone who heard them.
your answer
4 story prompts- Recall the last time you were pushed for a date you did not believe in.
- Write the range you would have given and the one condition that decides it.
- Name the smaller thing you could have guaranteed instead of the whole delivery.
- Check whether you ever recorded what you actually promised, and where.
draft and rehearse your own answer in a learn session
go deeper
This probes how you behave when honesty and pressure pull in opposite directions. The interviewer is testing whether you can give a partner something useful to plan against without inventing certainty, whether you understand why the date is being asked for, and whether the commitments you do make are ones people can rely on later.
Our founder asked me in front of the room for the date we would finish the logging cutover, because a customer agreement depended on the audit trail it fed. First I asked what the date was for, because a signature happening on a specific day and a general sense of whether this is a quarter or a year need different answers. It was the first. Then I gave the range and the reason it was one: six to eleven weeks, and the entire spread came down to a single broker change that the streaming team owned and had not scheduled. I said the six-week case assumed that change went into their next cycle. Instead of the date he wanted, I offered two things I could actually commit to. The audit-trail slice specifically, the part his customer needed, carved out and shipped independently of the rest of the cutover. And a hard date for a decision rather than a delivery: six days later, once we had talked to the streaming team, I would come back with either a firm date or a plan that did not need them at all. I wrote both down that afternoon so the promise on record was the one I had actually made. On the sixth day we came back with a firm eight weeks. The cutover week ran with two pages against the eleven a week the old pipeline had been throwing, and the customer slice landed before the agreement was signed.
What carries this is the substitution: refusing the delivery date but committing to a carved-out slice and a dated decision point, so the founder left with something real. Writing it down the same day is the small move that prevents the range being remembered as its optimistic end.
Once three groups were all asking my organisation for dates, I stopped answering the question one meeting at a time and changed what we hand over. We now publish two artefacts. The first is a commitment list: a deliberately small set of items with real dates, which something only joins after a de-risking investigation has closed its biggest unknown. Nothing goes on that list on the strength of a plan alone. The second is a forecast: everything else, as a range with a confidence and the named condition that would move it, refreshed on the first Monday of each month. So my conversation with an executive who wants a firm date is short now. I ask whether they are making an external promise on the back of it. If they are, we spend a week de-risking, and it moves onto the commitment list with a date I will defend. If they are not, they get the forecast, and I show them what buying down the unknown would cost in people and time. The hard part was internal. Three of my senior engineers had been giving optimistic point dates because it made meetings easier, and I had to make it genuinely safe to say 'I do not know yet, and here is when I will.' Two review cycles in, the paging load we forecast against came in inside the stated range for eleven of thirteen workstreams, and both misses had skipped the de-risking step. That is the argument I lead with now.
This works because the answer is a mechanism rather than a conversation: a standing separation between what earns a date and what stays a forecast, plus the internal culture work that made it stick. The evidence at the end measures the mechanism, not the speaker's own accuracy.
You mostly receive dates rather than issue them. Your version is telling your lead early that your own task has a range, what it depends on, and when you will know more — rather than nodding and hoping.
You are committing your own work. Give the range with its condition, and when reality moves, re-forecast in writing to everyone planning around you instead of absorbing the pressure quietly.
You commit on behalf of a team, so the alternatives you offer matter: a scoped slice you will guarantee, or a hard date for a decision rather than a delivery. Say where you recorded what you actually promised.
You change how dates get asked for. Talk about separating commitments from forecasts as standing practice, what earns a hard date, and how you made it safe for your engineers to say they do not know yet.
saying these in an interview costs you the question
- Agreeing to the date out loud while privately planning to slip
- Refusing to give any number at all on principle
- Casting the person asking as naive rather than hearing their planning need
- Offering a range so wide it carries no information
- Padding the number silently instead of naming the risk
- The range never narrows because nothing was scheduled to narrow it
- What if they escalate over your head?Show that you would rather be in that conversation than around it. Describe going to the escalation with the same range, the same condition and the same alternatives written down, and letting the decision be made with the information visible. Resentment or quiet compliance both read badly here.
- How wide is too wide for a range?Give a practical rule — a range so wide that both ends imply the same decision is useless to the listener. Say what you do at that point: investigate before quoting, or split off a slice you can commit to. Width should shrink as unknowns close, and you should say when.
- Have you ever committed to a date you did not believe in?Answer honestly, because most people have. Describe the pressure, what it cost, and what you now do at that moment instead. A candidate who claims they never have usually has not been close enough to a real commitment for the question to bite.