Three time zones rarely share a comfortable hour. The trap is treating the schedule as a negotiation, where the loudest calendar wins, instead of a small arithmetic problem with an honest answer. Here is the method that finds the real overlap, and what to do on the common days when there is no good overlap at all.
Step 1: put everyone on one clock
The mistake that wastes the most time is comparing local clocks directly. Nine in the morning is a different instant in every city, so a grid of local times hides the answer. Convert each person's working day to UTC first, then compare. A team that lives in San Francisco, London, and Singapore in June looks like this once you do that, with a standard nine to six working day:
- San Francisco (UTC-7 in summer): 16:00 to 01:00 UTC the next day.
- London (UTC+1 in summer): 08:00 to 17:00 UTC.
- Singapore (UTC+8, no daylight saving): 01:00 to 10:00 UTC.
You can do this conversion by hand, or paste the three cities into the time zone converter and read the offsets straight off.
Step 2: keep only the hours everyone shares
Now intersect the three UTC windows. London and Singapore overlap from 08:00 to 10:00 UTC, which is a tidy two-hour band: 09:00 to 11:00 in London, 16:00 to 18:00 in Singapore. Then add San Francisco, whose day does not begin until 16:00 UTC. The three-way overlap is empty. There is no hour of the day when all three are inside normal working time, and stretching San Francisco to a seven in the morning start does not fix it either.
This is the honest answer the local-time grid was hiding. The meeting scheduler computes this intersection for you and shades the hours that are humane, marginal, or brutal for each city, and the team time zone planner shows a whole team's day as a single strip so the gap is obvious at a glance.
A spread that does work, and why
Not every team is doomed to a call at 2am. The same arithmetic that exposes an impossible spread also finds the good slot when one exists. Take London, Berlin, and New York in summer, again on a nine to six day:
- New York (UTC-4): 13:00 to 22:00 UTC.
- London (UTC+1): 08:00 to 17:00 UTC.
- Berlin (UTC+2): 07:00 to 16:00 UTC.
Intersect them and a comfortable three-hour band survives: 13:00 to 16:00 UTC. That is 09:00 to 12:00 in New York, 14:00 to 17:00 in London, and 15:00 to 18:00 in Berlin. A late-morning call for New York, mid-afternoon for Europe, and nobody is awake at an unreasonable hour. The lesson is not that some teams are lucky, it is that the size of the spread, not anyone's preference, decides whether a humane slot exists. Around eight hours of offset difference is the practical ceiling for a daily live meeting that does not hurt someone. New York to Berlin is six hours and leaves a three-hour band; beyond roughly eight you are in rotation territory.
Step 3: when there is no overlap, rotate the pain
A spread this wide has no painless slot, so the only fair move is to share the cost. Pick the least-bad time, then rotate who carries it. One week the early call lands on San Francisco, the next on Singapore, so no single region is permanently handed the seven in the morning or the ten at night. Write the rotation down and put it in the calendar invite, because an unwritten rotation always drifts back to whoever complains least.
A fixed weekly slot feels simpler, but it quietly taxes the same people forever. That is the hidden reason global meetings get resented: not the meeting itself, but the same region absorbing the unsociable hour every single week.
A rotation that actually holds
Back to the impossible spread. With San Francisco, London, and Singapore there are really only two defensible slots, and the fair move is to alternate them week by week:
- Week one, 16:00 UTC. San Francisco 09:00 and London 17:00 are both workable; Singapore carries it at midnight.
- Week two, 09:00 UTC. London 10:00 and Singapore 17:00 are both workable; San Francisco carries it at 02:00.
London sits in the middle and is comfortable either week, which is a fair outcome rather than a flaw: the geography hands the overlap to whoever is between the extremes. The two outliers, San Francisco and Singapore, trade the unsociable hour evenly. Put both slots in the calendar as a fixed alternating pattern so nobody has to renegotiate it each week, and when a person is on the painful side, keep that meeting short and make the written notes excellent so missing it costs them nothing.
Step 4: make synchronous the exception, not the default
The cleanest way to win the scheduling problem is to have fewer live meetings. Most status sharing does not need everyone awake at once. Reserve synchronous time for the things that genuinely need it, a hard decision, a sensitive conversation, a kickoff where energy matters, and move the rest to a written, async channel. The async-first standup playbook covers how to do that without losing visibility, and which cases still justify the call.
Two things that quietly move the overlap
Daylight saving. The overlap you calculate in June is not the overlap in December. When one region shifts its clocks and another does not, the gap between them changes by an hour, and for a few weeks each spring and autumn the two sides of the Atlantic are briefly one hour closer or further apart than usual. A recurring invite stored as a fixed UTC offset will silently drift through these transitions. Store the meeting against an IANA time zone such as Europe/London rather than a frozen offset, and the calendar recomputes the correct local time on its own. The daylight saving survival guide lists every 2026 transition date and exactly how recurring invites break around them.
Half-hour and quarter-hour zones. India runs at UTC+5:30, Nepal at UTC+5:45, and central Australia at UTC+9:30. Round these to the nearest whole hour and your overlap is off by thirty or forty-five minutes, which is enough to turn a slot that looked fine into one that clips someone's lunch or the end of their day. Convert the exact offset rather than approximating it.
Common mistakes that recreate the problem
- Comparing local clocks directly. A grid of everyone's 09:00 hides the answer, because each 09:00 is a different instant. Convert to one reference first.
- Picking the average time zone. The midpoint between San Francisco and Singapore is the middle of the Pacific, where nobody works. Averages do not schedule meetings; overlaps do.
- Freezing one permanent slot. It feels tidy and quietly taxes the same region forever, which is how resentment builds without anyone naming it.
- Ignoring the transition weeks. Scheduling against the wrong side of a daylight saving change puts the call an hour off for half the team.
A repeatable checklist
- Convert every participant's working hours to UTC before comparing anything.
- Intersect the windows and keep only the hours shared by everyone.
- If the overlap is comfortable, book inside it and stop.
- If it is thin or empty, choose the least-bad slot and rotate who carries it.
- Move anything that does not need a live conversation to async.
Frequently asked questions
What is the overlap window for a meeting across time zones?
It is the set of clock hours that fall inside everyone's working day at the same instant. Convert each person's working hours to UTC, then keep only the hours that are inside all of them. Whatever survives is the humane window. Often, across three distant zones, nothing survives.
How do you schedule a meeting when there is no overlap?
Pick the least-bad slot and rotate it. If San Francisco, London, and Singapore cannot all meet inside business hours, take turns carrying the early or late call rather than always handing it to the same region, and default to async for anything that does not strictly need a live conversation.
Does a single fixed meeting time work for a global team?
Rarely, and never fairly. A fixed slot quietly assigns the same region the unsociable hour every week. A rotating slot, or an async-first default with occasional synchronous exceptions, spreads the cost and tends to last longer.
How wide a time zone spread can still share a daily meeting?
As a rule of thumb, about eight hours of offset difference is the ceiling for a daily live meeting that does not hurt anyone, because a standard nine-hour working day leaves only a thin overlap at that point. New York to Berlin is six hours and shares a comfortable three-hour band; San Francisco to Singapore is fifteen hours and shares nothing, so it has to rotate.
Should the people who are comfortable join the rotation too?
Only if a slot actually inconveniences them. In a wide spread the middle region often overlaps both extremes and is genuinely fine in every slot, so forcing it to take a bad hour purely for symmetry spreads pain without reducing it. Rotate the burden among the people a slot truly hurts, not as a token gesture.