Find the constraint, not the symptom
Bottlenecks, safety stock, rotas, landed cost, meeting cost and automation payback — with the assumptions shown. Nothing you enter leaves your device.
Capacity planner
Find the step in your process that everything else is waiting on.
Stages in your process
How this is calculated
For each stage: effective hours = (units × hours each − setup) × uptime × yield, then output = units per hour × effective hours. The stage with the lowest output sets the capacity of the whole line.
Real processes have buffers, variability and batching this doesn't model. Treat the bottleneck as the thing to investigate, not a measurement.
Staffing & shift planner
Check a rota covers what it needs to, and what it costs.
Shifts
People
List the shift numbers each person works, separated by commas — e.g. 1,3.
Inventory reorder point
When to reorder, and how much, so you don't run out while the supplier ships.
How this is calculated
reorder point = daily demand × lead time + z × √(demand variance × lead + demand² × lead variance)
The z value comes from your service level (95% → 1.65). It assumes demand and lead time are roughly bell-shaped and independent, which is a simplification — a supplier who is late precisely when everyone is busy breaks the assumption.
Landed cost & supplier comparison
What a unit actually costs once freight, duty, fees and rejects are counted.
Quotes
How this is calculated
landed cost per unit = (goods + freight + duties + fees) ÷ (units × (1 − defect rate))
Currency conversion, payment terms and the cost of holding a larger order are not modelled — a supplier who needs payment up front is more expensive than one who bills on delivery, even at the same price.
Meeting cost
What an hour in a room costs, including the preparation nobody counts.
Who attends
How this is calculated
cost = Σ(people × hourly rate) × (meeting + prep + follow-up hours), multiplied by how often it recurs.
This is the cost of the time, not a verdict. A weekly meeting costing $40,000 a year is a bargain if it prevents one bad decision, and appalling if it's a status update that could be a written note.
Automation ROI
Whether automating a manual task pays for itself, allowing for the fact that it never saves 100%.
How this is calculated
Annual manual cost is hours × loaded rate plus today's error cost. The saving is that figure × time saved × adoption, less ongoing maintenance and licences. Future years are discounted to today's money and netted against the upfront build.
Time freed is only a saving if it goes somewhere useful. Five minutes returned to twelve people is not one person's day back.
Operations is mostly finding the constraint
Almost every operational problem is one of these: something is the bottleneck and you're improving the wrong thing, stock runs out because the buffer was sized on a hunch, a supplier looks cheap until the rejects are counted, or time is spent on a task nobody has ever priced. These tools do that arithmetic and show their working, so you can argue with the assumptions rather than the answer.
Rotas, supplier prices and cost rates stay in your browser. Nothing is uploaded, stored or sent anywhere.