Three Measurements That Actually Run a Plant

Your controller reports green. Shipping is red. Every department hit its local KPI, yet customers wait and cash sits in WIP you cannot invoice. That gap is why Eliyahu M. Goldratt and Jeff Cox's The Goal (North River Press, 1984) replaces comfort metrics with three operational measurements every shift supervisor can use.

If you have not read the series intro, start there for context. This lesson translates headquarters finance into gemba language.

#The goal is money—not activity

Net profit, return on investment, and cash flow matter at headquarters. On the floor, they translate to throughput, inventory, and operating expense. Everything you manage fits one bucket. If a meeting debate cannot be mapped to one of the three, you are probably arguing about local efficiency instead of system performance.

The plant exists to make money. Quality, technology, and efficiency are means—not ends.

#Throughput: money in through sales

Throughput is the rate the system generates money through sales—not production. Parts sitting in a warehouse are not throughput until a customer pays. When a bottleneck machine runs three shifts building forecasted SKUs nobody ordered, you may be busy and still broke.

Throughput accounting (Goldratt's term) treats only sold product as money entering the system. Building ahead of demand converts cash into inventory—a balance-sheet move that does not improve throughput until someone signs a purchase order.

Ask at every decision: Does this increase what we can ship and collect for this month?

Tip. In your next ops review, replace "units produced" with "units shipped and invoiced." Watch how the conversation changes when production volume stops masquerading as success.

#Inventory: money stuck in the system

Inventory is money invested in things you intend to sell—raw material, WIP, finished goods, even tools and buildings you could sell. Goldratt deliberately excludes "value added" labor from inventory to stop managers from treating WIP piles as assets while they balloon carrying cost.

If inventory rises and throughput does not, you are tying up cash. A pallet of parts is not neutral—it burns cash daily through warehousing, insurance, obsolescence risk, and finance charges.

Controller Lou in The Goal notes all three measurements are defined in money terms—throughput in, inventory stuck, expense out. That simplicity is the point: one language from gemba to HQ.

#Operating expense: money out to convert inventory

All money spent turning inventory into throughput—payroll (direct and idle), utilities, scrap, carrying cost—is operating expense. There is no hiding labor inside inventory; time is expense.

Idle time at a non-bottleneck is operating expense. So is overtime at a non-bottleneck while the constraint waits for material. Both show up on the P&L. The difference is whether that expense buys throughput or buys WIP.

Carrying cost is real expense—even when it never appears on a department scorecard.

#The productivity test

An action is productive only if it increases throughput while reducing inventory and operating expense—or at least does not harm the other two. That test exposes "efficient" automation that lowered cost-per-part on paper while overdue shipments climbed and WIP choked aisles.

Most local hero projects fail this test. They optimize one operation's unit cost while flooding downstream with inventory nobody can assemble or sell this month.

#Throughput accounting vs cost accounting

When finance punishes inventory reduction on the balance sheet while operations free cash, teach the three measurements in the monthly ops review. Alignment prevents local heroes from rebuilding WIP for "asset" optics.

Traditional cost accounting allocates overhead across every operation and rewards lowering unit cost at each step. Throughput accounting asks a simpler question: Did the system make more money this month? Both frameworks have a place—but when they conflict, throughput thinking usually explains why shipping is red while every department is green.

Cost-world habit	Throughput-world counter
Maximize utilization everywhere	Maximize flow at the constraint
Treat WIP as a valued asset	Treat WIP as trapped cash
Cut expense at any operation	Cut expense that does not buy throughput
Reward cost-per-part reduction	Reward shipped-and-collected revenue

#Weekly ops review questions

Run these three questions every week until they become reflex:

1. Did throughput (shipped and invoiced) rise?
2. Where did inventory dollars increase without a matching sales increase?
3. What expense rose without a throughput gain?

One slide, three trends, twelve weeks. No utilization hero charts without throughput context.

#What to do this week

Walk one hot order from release to ship. Tag each wait as inventory, each labor hour as expense, each invoice line as throughput. Where do the three diverge from what your dashboards reward?

Bring one example to your leads meeting—a order that was "on track" in production but missed ship date because WIP sat between operations. That single story often opens the conversation better than any slide deck.

Tip. Give shift supervisors a laminated card: Throughput = shipped sales. Inventory = cash in system. OE = money to convert. Productive = helps all three.

#Shift supervisor card

Carry the three measurements into every handoff:

• Throughput — What shipped and invoiced since last shift?
• Inventory — Where did WIP rise or fall? Any pile building upstream of one operation?
• Operating expense — Any overtime, scrap, or idle time that did not buy throughput?

When a supervisor can answer all three in thirty seconds, you have a management system—not just a reporting system.

#Common measurement mistakes

These show up in nearly every plant the first time you apply throughput thinking:

Mistake 1 — Treating production volume as throughput. Units completed is not units sold. Forecast builds, safety stock, and "keep the line running" production all inflate activity without moving money in.

Mistake 2 — Ignoring carrying cost. Finance may capitalize WIP; operations still pays for floor space, handling, obsolescence, and the opportunity cost of cash locked in piles. Carrying cost belongs in operating expense thinking even when it never hits a department P&L.

Mistake 3 — Cutting labor at the constraint. Expense reduction at a non-bottleneck saves operating expense. Expense reduction that idles the bottleneck destroys more throughput than the savings are worth. Always ask: Is this resource the constraint?

Mistake 4 — Rewarding inventory builds. When buyers or planners get credit for " securing supply" by over-ordering, inventory rises without a matching throughput plan. Tie procurement incentives to consumption rate at the constraint, not to order volume.

Tip. In your monthly review, add one column: Cash impact this month. Throughput adds cash. Inventory growth subtracts. Expense subtracts. Three numbers, one cash story.

#Monthly plant review slide

Build one slide with three twelve-week trend lines—no utilization hero charts without throughput context:

1. Shipped and invoiced revenue (throughput proxy)
2. Total inventory dollars (material + WIP + FG)
3. Operating expense trend with one annotation for constraint overtime

When all three lines move in the right direction together, you have system improvement—not local heroics.

#Controller alignment workshop

Schedule one hour with finance before changing how production reports success. Walk through a single order that shipped late despite green production metrics. Tag each wait as inventory, each labor hour as expense, each invoice line as throughput.

Ask finance: Which of our current reports would have predicted this miss? The gap between operational reports and cash outcomes is where throughput accounting earns its keep.

When controller and plant manager share the three measurements in the monthly review, local heroes lose their cover. WIP builds for "asset" optics become visible as cash traps—not productivity wins.

#Building the habit on every shift

The three measurements only work when they become reflex—not a monthly report. Train every shift lead to answer three questions at handoff:

1. What shipped since my last shift? (throughput)
2. Where did WIP change? (inventory)
3. What did we spend that did not buy throughput? (operating expense)

When those three answers take thirty seconds, you have a management system. When they require a spreadsheet, you have a reporting system that arrives too late to change today's release decision.

#Tie to next lesson

Measurements mean little without flow. Local efficiency can destroy all three at once—pushing non-bottleneck machines to stay busy inflates inventory, raises carrying cost, and still misses ship date.

Next: The Flow Engine — why activating every machine is not the same as utilizing it.

← Start the series · Next: The Flow Engine →