Four hours of overlap is enough — if you spend it correctly
What to put inside the overlap window, and what should never happen there.
he objection to hiring in the Philippines, when it is not about cost, is almost always about hours. The team is twelve or fifteen hours away; what can four hours of shared time possibly be good for? The answer is: everything that needs two people in the same moment, which is a much shorter list than most teams behave as though it is.
Philippine time is UTC+8, which makes the arithmetic easy. Sydney is two hours ahead, so a normal Manila day already overlaps most of an Australian one. US East is twelve hours away — a Manila evening from seven to eleven is a New York morning from seven to eleven. US Pacific is fifteen, which is the case where someone has to work genuinely unsociable hours and should be paid a night differential for it.
What the window is for
Four things, and they are the only four.
Decisions that need a room. Not the decision itself — that can be written down — but the twenty minutes of disagreement that produces it. Threads are a bad medium for disagreement; they make people polite and slow.
Unblocking. Someone is stuck on a thing where the knowledge lives in a colleague's head and not in any document. This is the highest-value use of shared time there is, and it is the first thing a status-heavy standup crowds out.
Pairing on the hard part. Not the whole task. The one piece where two people looking at the same screen will finish in forty minutes what one person would take a day to work out alone.
Anything with weight. Feedback, disagreement about scope, bad news, a change of plan. Delivering these asynchronously reads as evasion even when it is only logistics.
What should never happen there
Status. Demos that could be recorded. Reading out a document everyone could have read. Any meeting whose output is a decision one person was always going to make alone.
The test is simple and it is worth applying ruthlessly: if this meeting would produce the same artefact had it been a document with comments, it should have been a document with comments. Overlap is the scarcest resource a distributed team has, and the default behaviour of every organisation is to spend scarce resources on ceremony.
A team that writes well needs less overlap than a team that does not. That is not a slogan; it is the mechanism. Every decision captured in a document is a decision nobody has to stay up to hear.
Overlap has a cost, and it lands on someone
This is where the research is useful. A CHI 2023 study by Mok, Sun, Sen and Sarrafzadeh analysed twenty million meetings at Microsoft and found that cross-time-zone meetings cluster into early-morning and late-evening hours — against employees' own stated preferences — and that the burden of meeting across time boundaries falls asymmetrically depending on where you sit.
That asymmetry does not resolve itself. Left alone, the overlap drifts to whatever is convenient for the largest or most senior group, and the cost is quietly transferred to the people furthest from headquarters. The team that is one hire deep in Manila absorbs it, says nothing, and the first signal you get is a resignation.
So make it explicit. Decide where the window sits and why, write down who is giving up an evening and who is giving up a morning, and rotate anything that cannot be made symmetrical. If the honest answer is that the Philippine team will carry the unsociable hours permanently, then that is a US Pacific shift pattern, and it is priced accordingly — a conversation about money is a far better outcome than an unspoken expectation.
An analysis of 20 million meetings at Microsoft finding that cross-time-zone meetings cluster into early-morning and late-evening hours, against employees’ stated preferences.