Why This Matters More Than It Seems

A few minutes here and there sounds trivial.  It is not.  On a single line running two shifts, 30 minutes of unrecorded lost time a day adds up to more than 100 hours a year.  That is capacity you are already paying for in wages, power and equipment, and not getting back as product.

Worse, because it does not show up in the records, nobody is tasked with fixing it.  It becomes part of the background, and the business plans around a level of output it cannot reliably achieve.

It also shapes the big decisions.  A business that believes its line is running flat out will start pricing a second shift, a new machine or a bigger site.  Those are serious commitments.  Before making any of them, it is worth knowing whether the capacity you already have is being fully used, because in many cases it is not.

Pattern 1: The Run Starts When the Machine Starts, Not When Good Product Appears

Ask an operator when a run started, and most will give you the time they switched the machine on.  It is a reasonable answer.  It is also usually wrong, in a way that matters.

On most lines, there is a gap between machine-on and the first good unit coming off the end.  Settings need checking.  The first few units are run as tests and scrapped.  Material needs to come up to temperature, quality checks need to pass, and someone walks the length of the line to make sure everything is right before committing to the run.

None of that is wasted effort.  It is part of doing the job properly.  But when the record shows the run starting at machine-on, all of that setup time is quietly counted as production time.

What it does to your numbers

If the record says the line ran for seven hours and produced a certain volume, your output rate is calculated over seven hours.  If 20 minutes of that was setup and test units, the line actually produced that volume in six hours and 40 minutes.

That sounds like a small correction.  The problem is what happens next.  Management looks at the output rate, compares it to the machine’s rated speed, and concludes the line is running slowly.  The response is often to push for speed, which the line usually cannot deliver, rather than to look at how long it takes to get each run up and stable.

You end up chasing the wrong problem, with numbers that point you there.

Why it goes unnoticed

This pattern survives because nobody is being dishonest.  The operator writes down the time they started working on the run, which is a perfectly sensible interpretation of the question.  The form does not ask for anything more specific, so the distinction between “started” and “producing” never gets captured.

When I work through this on the floor, the gap is often raised independently by more than one person once you ask the right question.  Operators and supervisors usually know it exists.  It just never made it onto the form.

The fix is small

Add one field: time of first good unit.  Keep machine-on as well.  The difference between the two becomes one of the most useful numbers in the business, because it shows you exactly how much time each run spends getting ready, and which products, shifts or crews take longest to get there.

Pattern 2: Changeovers and Stops Take Longer Than the Record Says

The second pattern is closely related.  When I time changeovers on the floor and compare them with what has been written down, the recorded figure is almost always shorter.  Not by a few seconds, but often by a third or more.

It happens for the same reason as the start-time gap.  The person recording the changeover writes down the part they think of as “the changeover”: swapping the material, changing the tooling, adjusting the settings.  What does not get counted is everything either side of it.

The time that falls between the cracks

A typical changeover has a tail at both ends.  Before it, there is the wait for material to arrive, for a forklift or crane to be free, or for someone with the right skills to come over.  After it, there is the restart: test units, checks, small adjustments and the slow ramp back up to normal running.

Each of those feels like something other than the changeover, so each gets left out.  Add them up and a changeover recorded as 20 minutes can easily take 35 from the last good unit of one run to the first good unit of the next.

The same thing happens with unplanned stops.  A short stop to clear a jam or wait for material often is not recorded at all, because it was “only a few minutes”.  Ten of those across a shift is the better part of an hour.

Breaks are rarely as long as the roster says

Scheduled breaks are another source of drift.  The roster might allow 30 minutes, but the line rarely stops exactly when the break starts and restarts exactly when it ends.  It slows down before the break as people wind up what they are doing, and it takes time to get going again afterwards.

When I compare the gap between the last unit before a break and the first unit after it, the real loss is frequently well beyond the scheduled time.  Across a day, that can add up to a meaningful block of lost production that nobody has planned for and nobody is tracking.

Why changeovers deserve special attention

Changeovers matter more than most stops because they are frequent and predictable.  That makes them one of the easiest places to recover capacity.  Grouping similar products to reduce the number of changeovers, staging materials before the line stops, and having a clear standard method for the changeover itself can each recover meaningful time.

But you cannot improve what you cannot see.  If the record says a changeover takes 20 minutes, nobody is going to prioritise reducing it.  If the stopwatch says 35, it becomes an obvious target.

Pattern 3: The Most Important Field Is the One Left Blank

One of the first things I do when reviewing an operation is collect a month or so of production records and transcribe them into a spreadsheet.  It is slow, unglamorous work, and it almost always turns up something nobody expected.

The most common surprise is a field that is never filled in.  Often it is the one that matters most: actual output.  The form has a box for it, the box has been there for years, and on form after form it is empty.

How it happens

Usually there is a simple explanation.  The output figure is recorded somewhere else, perhaps in a system the office updates later, so the operator does not see the point in writing it twice.  Or the total is not known until the run is fully counted, by which time the form has already been handed in.  Or the field was added for a reason that has long since been forgotten, and nobody ever followed up when it stopped being used.

Whatever the reason, the consequence is the same.  The document that is supposed to connect time on the line with product made cannot do its job.  You can see when the line started, when it stopped and why it stopped, but you cannot calculate how much it produced per hour of running from that same source.

Why that matters

When output and time are recorded in different places, by different people, at different points in the day, they rarely line up cleanly.  That makes it very hard to answer basic questions with confidence:

• Which products run efficiently, and which do not?
• Is one shift or crew consistently more productive than another?
• Did last month’s improvement actually change output, or did it just feel better?

Most businesses I work with can tell me what they shipped last month.  Far fewer can tell me, from their own records, what each hour of line time produced.

Look at the forms themselves

The lesson here is simple.  Do not assume a form is being completed because it exists.  Pull a stack of recent ones and check, field by field, what is actually being filled in.  Blank fields, fields filled with the same number every day, and fields completed in obviously different handwriting at the end of the shift are all signs the record is not doing what you think it is.

It costs nothing to check, and what you find will tell you how much confidence to place in every report built on top of it.

Pattern 4: Chasing Speed When Run Hours Are the Real Lever

When output falls short of target, the instinct is to look at speed.  Can the line run faster?  Can the team work harder?  Is the equipment up to the job?

Sometimes those are the right questions.  More often, the line is already running close to the speed the process allows.  Many processes are constrained by physics, material or quality, and simply turning the dial up produces more scrap, more stoppages or more wear on the machine.

Output is speed multiplied by time

Daily output comes down to two things: how fast the line runs while it is running, and how many hours it actually runs.  If speed is fixed by the process, the only lever left is time.

That changes the question.  Instead of asking “how do we go faster?”, the business should be asking “how many hours a day is this line actually producing good product, and where does the rest go?”

A simple illustration: a line is available for ten hours a day.  Between setup, changeovers, unplanned stops and breaks that run over, it actually produces for eight.  Recovering just 30 minutes a day lifts output by around 6%, with no new equipment, no extra shift and no pressure on the team to work faster.

Why speed gets the blame

Speed gets blamed because of the three patterns above.  When setup is counted as running time, changeovers are under-recorded and output is not captured alongside time, the records show a line that is running for most of the day but producing less than it should.  The natural conclusion is that it must be running too slowly.

Fix the records, and the picture usually changes.  The line turns out to run at a reasonable speed when it is running.  The problem is how much of the day it spends not running, and that is a far more solvable problem.

I wrote about a related trap in the four manufacturing bottlenecks that look like capacity problems.  The common thread is the same: what looks like a shortage of capacity is often capacity the business already has but cannot see.

Where recovered time usually comes from

In my experience, the biggest gains rarely come from one big fix.  They come from several small ones:

• Getting each run to its first good unit faster, with clear setup standards
• Grouping similar products to reduce the number of changeovers
• Staging materials and equipment before the line stops, so changeovers are not waiting on anything
• Planning breaks so the line winds down and restarts cleanly, or covering them where it makes sense
• Making sure the people who handle stops have what they need to restart quickly

None of these is complicated.  But none of them gets prioritised until the business can see, in its own numbers, where the time is going.