Takt time is a demand number, not a target

Takt time gets printed on a board at the end of a line and then treated as a performance target, which is roughly like printing the tide table and treating it as a swimming goal. It is not something the line achieves or fails. It is a statement about the customer, and the line had no part in setting it.

The word is German, for a beat or a bar of music, and the metaphor is the whole idea: the customer sets the tempo and the line plays to it. Get that the wrong way round and everything downstream of it goes wrong too.

The number

Two inputs, and only two. How much time is genuinely available for production, and how many units the customer wants in it.

Both are more contested than they look. Available time is not shift length: a line that stops for breaks and a clean down does not have those minutes, and counting them produces a takt the line can never hold and a permanent, unexplained shortfall.

---

Takt = available time ÷ demand

	Shift length, eight hours	480 min
	Less two breaks of ten minutes	20 min
	Less start up and clean down	10 min
	Available production time	450 min
	Customer demand, 1,350 a week over five days	270 a day
	Takt time, the beat the customer is setting	100 seconds

Illustrative figures. Both inputs come from things a plant already knows: the shift pattern, and the order book. Neither needs a study.

A hundred seconds. That is the interval at which a finished unit has to leave the cell for the customer to be satisfied. It says nothing about whether the cell can do it.

What it is actually for

Here is the use that justifies calculating it, and it is not the board at the end of the line.

Add up the manual work content across the whole cell: every operation, at every station, for one unit. Suppose it comes to nine and a half minutes, which is 570 seconds. Divide by takt.

---

Operators = work content ÷ takt

	Total manual work content for one unit	570 s
	Takt time, from above	100 s
	Theoretical minimum operators	5.7
	Operators the cell needs	6

Five point seven is not a staffing level, it is a floor. Nobody achieves a perfect balance, so six is the honest answer and the 0.3 of an operator is the balance loss the cell will have to carry.

That is the answer takt exists to give. How many people, at what stations, doing what work. It is a design calculation, and it belongs on the day the cell is laid out and on every day the demand changes, not on a whiteboard where the shift is scored against it.

When the line cannot hold it

Now suppose the slowest station in that cell takes 118 seconds. The cell’s cycle time is 118 seconds, takt is 100, and the gap is where the argument starts.

Over a day of 270 units, 118 seconds each is 31,860 seconds of work against 27,000 seconds available. The cell is short by 4,860 seconds, which is 81 minutes. There are not 81 minutes left in the shift, so this is not a problem that effort closes.

There are exactly four honest answers, and pretending there is a fifth is where the whiteboard comes from.

1. Take eighteen seconds out of the slowest station. Rebalance the work across the six operators, or improve the operation itself. This is the real project, and it is the one worth doing.
2. Add time. Overtime, or a second shift. Available time goes up, takt goes up with it, and the cell holds.
3. Add a station. Which means the work content divides differently and the minimum operator count changes.
4. Supply less than the customer wants. Which is a decision, and it should be made deliberately by somebody entitled to make it rather than by a shortfall nobody discusses.

Telling the operators the number is 100 and asking them to try harder is not on the list. They already know, they can read the same board, and the work content is what it is.

The part that gets left out

A cell balanced to exactly takt has no room for anything to go wrong.

Follow the arithmetic. If every station is set at 100 seconds and the day’s available time is 450 minutes for 270 units, then the day contains no spare time at all. One ten minute stoppage is six units short, and there is nowhere in the shift to make them back. Every minute lost is lost permanently.

Which is why a balance target is set below takt rather than at it, and why the size of that gap is a decision rather than a convention. It is a judgment about how much variation the line genuinely has: how often it stops, for how long, and how expensive a shortfall is. A line that stops twice a shift needs a different margin from one that stops twice a month, and neither of them can find their number in a textbook. What both of them can do is measure their own stoppages for a fortnight and set the margin from that.

And it moves

The last thing worth saying about takt is that it is not a constant, and treating it as one is how a cell laid out correctly in March is wrong by June.

Demand changes. Run the same cell against a weekly order book that swings between 1,050 and 1,650 units and the takt swings with it:

At 129 seconds the cell needs 570 divided by 129, which is 4.4, so five operators. At 82 seconds it needs 7.0, so seven. The same cell, the same work, three different staffing answers, and rebalancing a line is not free.

So the useful discipline is to compute takt against a demand window you can actually hold to, and to treat the window as a decision rather than a fact. Weekly takt on a line that cannot be rebalanced weekly is a number that will be wrong more often than right. Monthly or quarterly takt, with a deliberate margin for the swing inside it, is a number a cell can be built around.

Takt, cycle time and lead time are three separate quantities and they get used interchangeably in the same meeting more often than any other trio in this field. Takt is what the customer needs. Cycle time is what the cell can do. Lead time is what an individual order experiences from the moment it is placed, and it is usually the largest of the three by an order of magnitude, for reasons that have nothing to do with the cell at all. That gap is worth its own piece, and it has one here.