Tag: Continuous Improvement

You now have a number: what one hour on your constraint is worth. Here is how to spend it this week. None of these require a project, a budget, or anyone’s approval. 

1. Should we take the line down for the planned work? 

This argument usually runs on instinct and organizational politics. Operations does not want the downtime. Maintenance does not want the failure. Both positions are reasonable and neither is quantified, so the loudest voice wins and the decision gets remade every month. 

Reprice it. 

A four hour planned intervention on our packaging line costs $24,000 in contribution margin. That is the cost of saying yes. 

Now the other side. Pull the history on the failure mode you are trying to prevent. If the unplanned version has historically run eight hours, it costs $48,000 each time it occurs. If it has occurred twice in the last twelve months, the expected annual exposure is $96,000. 

The comparison is now arithmetic, and it can be documented, revisited, and defended. Sometimes the answer will be to defer, and that is fine. A deferral you can justify is worth more than a shutdown you cannot. 

2. Is the spare part worth stocking? 

Inventory arguments are among the longest running and least resolvable disputes in most plants, because the two sides are measuring different things. Materials management is measuring carrying cost and working capital. Maintenance is measuring the wait. 

Price the wait. 

A critical part with a six hour lead time on a constraint asset is protecting $36,000 every time the failure occurs. A $9,000 part with a two year expected life, against a failure mode that occurs annually, is not an inventory decision at that point. It is obviously correct, and it takes one line to show. 

The reverse also applies, and you should be willing to run it. Plenty of parts sitting in stores protect assets with protective capacity, where the wait costs nothing. Pricing the hour cuts both ways, which is exactly what makes it credible when you use it in your favor. 

3. Was the overtime worth it? 

Every plant approves recovery overtime and almost none of them measure the return. 

An overtime shift that genuinely recovers five constraint hours returns $30,000 against a few thousand dollars of premium labor. That is a decision worth repeating and worth telling your plant manager about. 

An overtime shift spent on an asset that was never the constraint returns nothing, regardless of how much work got done. That is also worth knowing, and the honest version of this analysis is what makes people trust the favorable version later. 

Run last month’s overtime through the rate. Whatever the answer is, you will have learned something your plant did not previously know. 

The pattern 

All three of these are decisions your plant is already making, repeatedly, using judgment. None of them are bad decisions. They are simply unpriced, and unpriced decisions drift toward whoever argues most persistently rather than toward whatever is worth the most. 

You are not adding a process. You are adding a number to a conversation that was already happening. 

This week 

Take one real decision from your own plant this month and reprice it with your rate. Write down the before and the after. 

Bring that repriced decision into next week, because next week we stop estimating and start counting, and you will want an example of what the rate is for.  

Once you have sized your hidden plant, the first challenge you get will be about the denominator. It is a fair challenge and you should have the answer ready, because the two available answers are both correct and they serve different audiences. 

Against scheduled time 

This measures the hours you staffed, powered, and paid for. Weekends you did not run, shifts you did not staff, and planned shutdowns are excluded. 

It is the plant manager’s number. It is fair to the crew, because it only counts time the plant asked the asset to produce. It is defensible on the floor, because nobody can argue that you are charging them for hours the business chose not to schedule. And it is the right basis for the case you are going to build, because the action you will propose operates inside scheduled time. 

On our packaging line, that number is $10.5mm a year. 

Against calendar time 

This measures the asset you actually bought, all 8,760 hours of it. 

It is the capital committee’s number. It answers a different question: how much of the equipment on the balance sheet is producing margin, and how much capacity is available without buying new steel. It is the number a private equity buyer runs during diligence, and it is the number that competes directly with a capital request for additional capacity. 

It is always larger, sometimes dramatically so, and it is uncomfortable in a way that is occasionally productive. 

Why the choice matters more than the arithmetic 

Use the wrong one in the wrong room and you lose the room. 

Put the calendar time figure in front of a crew that has been working hard and you have told them, in effect, that their best week was a fraction of what it should have been. It reads as an accusation regardless of your intent, and the honest counting you are about to ask them for will not happen. 

Put the scheduled time figure in front of a capital committee that is weighing a new line and you have understated your own case, because the committee is deciding about the asset, not about the shift schedule. 

The rule is simple. Build the case on scheduled time. Keep the calendar time number in your pocket for the capital conversation, and produce it only when someone proposes buying capacity you may already own. 

The reflex this is aimed at 

When demand rises, the reflex is a capital request. New line, new packer, new building. The business case is straightforward, the approval path is well worn, and the request moves quickly because everyone involved knows how to evaluate it. 

Meanwhile the hidden plant sits inside assets already on the balance sheet. It requires no capital approval, no installation window, no ramp up curve, and no additional floor space. And it produces nothing, because nobody has sized it and therefore nobody has proposed it. 

The question worth putting in front of leadership is not whether to add capacity. It is whether to collect the capacity already purchased before purchasing more. That question has never been asked in most plants, and it is not asked because the number required to ask it does not exist. 

The objection you will get 

“We cannot sell everything we make.” 

This is the right challenge and it deserves a real answer, which is that recovered hours can be taken three ways. As volume, where they convert to margin at full value. As cost, by running fewer scheduled shifts for the same output and removing premium labor and utility hours from the base. Or as capacity held, where the hidden plant becomes a capital avoidance argument and the next line gets deferred by years. 

The mistake is treating an unsellable hour as a free one. It is not free. It was paid for. 

This week 

Calculate both numbers. Present one. Know which room you are in. 

There is a plant somewhere this month that will beat plan by three percent, hold a short celebration, and leave $10mm of contribution margin on the floor. Nobody involved will do anything wrong. 

The mechanism is the plan itself. 

How the plan absorbs the losses 

Production plans get built from history, because history is the most defensible input available to a planner. Last year’s actual output becomes this year’s baseline, adjusted for demand, mix, and known changes. Nobody would design it differently, and as a scheduling instrument it works. 

The problem is what history contains. Last year’s actuals already include last year’s unplanned stops, last year’s slow running, last year’s long changeovers, and last year’s rework. All of it is priced into the baseline as if it were a property of the asset rather than a set of losses that could be recovered. 

So the losses get inherited, and then they get hidden, because once they are inside the plan they stop being losses and start being the plan. 

Beat it and you are performing. Miss it and you are underperforming. In neither case does anyone ask the question that finds the money: what was the asset capable of? 

The two plants 

It is easier to hold if you think of it as two plants operating in the same building. 

The visible plant is what you shipped. It is measured, reported, budgeted, forecast, and rewarded. Everything about it is well governed, and it is the plant your P&L describes. 

The hidden plant is the output you already have every resource to produce and are not producing. Fully staffed, fully powered, fully supplied, fully paid for, and not collected. It has no reporting, no owner, and no line in the budget. 

The hidden margin is the money trapped in the second plant. It is contribution margin that the fixed cost base has already been paid to produce. 

That last point is what makes this an executive conversation rather than a maintenance one. You are not proposing to spend money to create capacity. You are proposing to collect capacity you have already bought. 

What it looks like in numbers 

Back to the packaging line. Case packer constraint, 600 bags per hour design rate, 417 scheduled hours in the month, $10 contribution margin per bag. 

Design output at scheduled time: 250,000 bags. 

Actual good output: 162,500 bags. 

Hidden plant: 87,500 bags per month. 

At $10 per bag, that is $875,000 a month, or $10.5mm a year, on one asset. 

Last week the same asset produced a $1.5mm figure when measured against plan. The plan was concealing a factor of seven. 

Note what this figure is measured against. Scheduled time only. It excludes weekends and unstaffed shifts entirely, which is what makes it fair to the crew and defensible on the floor. 

The word that matters is hidden 

Not lost. Not wasted. Not broken. 

Hidden means already owned and not collected, and that is a fundamentally different conversation to have with a CFO. Lost capacity sounds like an accusation. Uncollected capacity sounds like an asset, which is exactly what it is, sitting on a balance sheet you are already depreciating. 

This week 

Take your asset. Find its design rate, confirm it against the best sustained rate you have documented evidence for, and use the lower figure. Count scheduled hours for one representative month. Multiply. Subtract actual good output. Multiply the gap by contribution margin. 

Then resist the urge to soften the answer. 

This is the whole calculation. It takes longer to schedule the meeting about it than to do it. 

The four inputs 

1. Design rate. Units per hour the asset was built to produce. Start with the nameplate or the original equipment documentation. 

2. Scheduled hours. Hours in a representative month that the plant asked this asset to produce. Exclude unstaffed shifts and planned shutdowns. 

3. Actual good output. Units in that same month that were saleable. Good output, not gross output. Anything reworked or scrapped consumed a constraint hour you cannot resell. 

4. Contribution margin per unit. Price less variable cost, sourced from finance in writing. 

The arithmetic 

Design rate times scheduled hours gives design output. 

Design output less actual good output gives the hidden plant in units. 

Hidden plant units times contribution margin gives your hidden margin. 

For our packaging line: 600 bags per hour times 417 hours is 250,000 bags. Less 162,500 actual gives 87,500 bags. At $10 per bag, $875,000 a month, $10.5mm a year, one asset. 

When the design rate is a problem 

This is where most people stall, so here is how to handle each case. 

The nameplate is missing. Ask the original equipment manufacturer, who will usually have it against the serial number. Failing that, use the best sustained rate the asset has actually demonstrated, documented from a production log rather than from memory. 

The nameplate is obviously inflated. Common, particularly where the asset was specified for a different product or package format than it now runs. Use the best sustained demonstrated rate instead and note the substitution in your assumptions. 

Nobody agrees on it. Ask two people who have run the asset at its best what it does on a good day. Take the lower figure. You are not trying to win an argument about the ceiling. You are trying to establish a number that survives challenge, and the conservative version does that better. 

The principle throughout: a defensible number you can hold beats an ambitious number you have to retreat from. Every point you concede on the design rate makes the remaining figure harder to dismiss. 

Write down your assumptions as you go 

Four lines is enough. 

Design rate, and where it came from. Scheduled hours, and what you excluded. Actual good output, and whether it is good or gross. Contribution margin, and who at finance provided it. 

Those four lines are what turn your figure from an opinion into a calculation. The first person who challenges the number will challenge one of them, and having the answer ready is what ends the challenge rather than starting a debate. 

Expect it to feel too big 

It will. Nearly everyone’s first reaction to their own hidden plant figure is that it must be wrong, because a number that size would surely have been noticed. 

It has not been noticed because nothing in the reporting system was built to notice it. The P&L cannot see uncollected capacity. The plan has already absorbed the losses. OEE, where it is tracked at all, is frequently reported as a percentage without ever being converted into money. 

Do not shrink the number to make it comfortable. Test it, which is exactly what the next two weeks are for. If your counted losses land within about ten percent of your calculated gap, the number was right. 

Your twenty minutes 

Run the four inputs. Write the four assumption lines. Put the result in one sentence: this asset gives away X dollars of contribution margin per year at current performance. 

Then carry that sentence into next week, where we start finding out where the hours actually went.