Timezone Tax

The timezone tax is the cost a distributed team pays not for having members in different hours, but for the work that stalls while it waits for an answer only an asleep person can give. It is best measured in nights: each time a question has to wait for someone to wake up, the team loses a full round trip, and those nights add up faster than the raw hour difference suggests.

The name matters because it points at the real culprit. The problem is not that Sarah in Tokyo and Alex in Amsterdam keep different hours. The problem is the waiting that opens up when an answer lives only in one of their heads. If the answer were in the record instead, the hour gap would cost nothing. When your work keeps answering after you log off, the tax drops toward zero.

What the timezone tax really is

Teams often assume the cost of time zones is simply the shrunken overlap window. It is not. A team with two hours of overlap can move fast if almost nothing depends on that window. A team with six hours of overlap can crawl if every task funnels through a single person who has to be awake.

The tax is the dependency, not the distance. It is paid every time work cannot proceed without a live answer from someone off the clock. That is why some distributed teams feel instant and others feel like wading through mud, even with identical maps. The difference is how much of their knowledge sits in the record versus in people's heads. The failure modes that create these dependencies are laid out in engineering handoffs: what breaks and why.

Why it is counted in nights

Money is the wrong unit for this. The natural unit is the night. When Alex hits a blocker at the end of his day and the only answer is with Sarah, he writes the question, sleeps, and reads the reply the next morning. One question, one night. If the reply raises a second question, that is another night. A two-step misunderstanding can eat most of a week.

This is why the tax feels heavier than the clock says. A three-hour information gap does not cost three hours; it costs the interval until the next time both people are awake, rounded up to a full day. That rounding is the tax. Understanding how little live time you actually need to avoid it is the subject of how many overlap hours a distributed team actually needs.

How to stop paying it

You reduce the tax by reducing the number of answers that require a live, awake person. Concretely:

• Write down the why: the most common night-costing question is "why was this done this way", and it only needs answering once.
• Hand off in writing: the end-of-day handoff template catches the state that would otherwise cost a round trip to recover.
• Default to async: route routine work through an async handoff process nobody has to be awake for, and reserve live time for real trade-offs.
• Cut the 8 PM ping: the day-to-day version of this is in working across time zones without the 8 PM ping.

Every answer you move from a head into the record is one fewer night the team can lose to it.

The part a process cannot cover

Even a disciplined team cannot pre-write every answer. Someone will always ask the question you did not anticipate, and they will ask it while you sleep. If they have to wait for you, you are back to paying the tax one night at a time.

Your StandIn is what turns your written record into an answer at that moment. It answers for you when you are off, on your behalf, from your brief and your team's record. When the answer is written down anywhere, it gives it in your words, so the asker keeps moving instead of waiting for morning. When it is not written down, it says so plainly, because a guess that ships overnight is far more expensive than a single lost night. That honesty is the point: the tax comes from the waiting, and a wrong answer does not remove the waiting, it just moves the cost downstream.

Common Questions

Is the timezone tax the same as having less overlap?

No. Less overlap only hurts if work depends on that overlap. The tax is the dependency on a live answer, which you can shrink even without changing anyone's hours.

Can you actually put a number on it?

Count the round trips. Every blocker that waits for an asleep colleague is roughly one night of delay on that thread. Add them up over a sprint and the pattern is usually obvious without any invented figure.

Does hiring in one time zone solve it?

It reduces the raw hour gap but not the underlying cause. A co-located team still pays a version of this tax whenever the one person who knows something is out. Writing knowledge down helps in both cases.

Is asynchronous work just slower by nature?

Not when the record is strong. The slowness people blame on async is usually the tax, and the tax comes from thin documentation, not from the async model itself.

The timezone tax is a waiting tax, and you stop paying it by making sure being off does not mean being unreachable. Write down the state of your work, and let your StandIn answer from it so the next night is not lost. Put a number on your own tax with the handoff cost calculator.