Key takeaways
Lot sizing is one of the most consequential planning decisions because it sits at the center of the trade-off between efficiency and flexibility. Make too much at once and you drown in inventory; make too little and changeovers eat your capacity.
Every order carries two opposing costs. Setup or ordering cost is incurred each time you run, so fewer, larger lots reduce it. Holding cost grows with inventory, so smaller, more frequent lots reduce it. Lot sizing is the search for the quantity that balances the two.
When changeovers are cheap and fast, small lots win and you move toward one-piece flow. When changeovers are long and costly, larger lots protect capacity, which is exactly why reducing setup time unlocks smaller lots.
A part has annual demand of 12,000 units, a setup cost of 200 per run, and a holding cost of 3 per unit per year. EOQ works out to about 1,265 units per run, roughly ten runs a year.
Order in lots of 6,000 instead and you halve setups but more than double average inventory. Order lot-for-lot at 1,000 per month and you cut inventory further but add setups that the bottleneck may not have time for.
The math points to a balance, but real capacity decides whether you can afford the changeovers that small lots imply.
Lot sizing assumes you can absorb the changeovers it implies. If the line is already capacity-constrained, every extra setup steals run time, and low OEE makes that worse. Measuring real availability shows how much changeover the schedule can actually take, so lot sizes match reality.
Book a Fabrico demo to see how live OEE data informs batch decisions. Lot sizing also feeds the schedule itself; see APS and finite-capacity scheduling .
Lot-for-lot makes exactly what each period needs, minimizing inventory but maximizing setups. EOQ picks a fixed economic quantity that balances setup and holding cost, best for steady demand.
Cheaper, faster setups lower the cost of running often, which shifts the economic lot size downward and makes small-batch flow affordable.