There’s a version of every manufacturing business where the shop floor and the board are describing different companies. The floor knows which machines are unreliable, which products fight them, and which customers cause chaos. The board sees margin by product line and a service percentage. Both sets of information are true; neither reaches the other in a usable form.
Where The Disconnect Costs Money
Pricing decisions made without real cost data. If your standard costs use theoretical cycle times, products that consistently run slow are quietly subsidised by ones that run fast. Sales pushes the wrong products and the margin mix deteriorates — with everyone acting on the information they were given.
Investment decisions made from the wrong end. Capital gets allocated on capacity assumptions rather than on demonstrated constraint data, which is how a plant ends up with a new machine at a station that was never limiting output.
Customer commitments made blind. Sales quotes lead times from a table. The floor knows that product hasn’t hit that lead time in a year.
The Customer Finds Out Last.
And improvement effort aimed at the wrong problems, because the loudest issue gets attention rather than the costliest one.
What Needs to Travel Upward
Less than people assume. Four things carry most of the value.
Actual versus standard run times by product and operation, which turns costing from an estimate into a measurement.
Downtime by cause, categorised consistently, which tells you where capacity is genuinely being lost.
Scrap and rework by product and stage, at full cost rather than material cost.
Schedule adherence with reason codes for misses — the single best leading indicator of delivery problems.
And What Needs to Travel Down
This direction gets neglected and matters just as much. The floor should see the demand picture, the customer commitments behind today’s schedule, and the cost consequences of what happens on their line.
Teams make better decisions with context. An operator who knows a job is for a customer with a hard deadline sequences differently from one who sees a job number. This isn’t motivational messaging — it’s giving people the information their decisions require.
Making Capture Realistic
Data quality from the floor depends almost entirely on how much effort collection takes. If recording a downtime reason requires walking to a terminal and navigating four screens, you’ll get generic codes entered at end of shift from memory.
Keep the reason code list short — eight to twelve categories — and make entry fast and local. Then, critically, feed the results back. Teams asked to record data that never returns to them in any visible form stop taking it seriously, and the decay is quick.
The Test
Ask a supervisor what the three biggest sources of lost output were last month. Then ask a director the same question. If the answers differ substantially, the connection isn’t working — and the supervisor is more likely to be right.
ticktick.ai captures operation-level production data and surfaces it in both directions — cost and constraint views upward, demand and commitment context downward.
