Smaller Batches, Faster Flow: A Practical Guide to Reducing Batch Size
Large batches feel efficient. You process a hundred invoices at once, run a big production lot, or hold work until you have 'enough' to make a meeting worthwhile. It feels like you are saving on setup. In Lean thinking, though, big batches are one of the most reliable ways to slow a process down, hide defects, and tie up cash. Reducing batch size is one of the highest-leverage moves available, and the disruption of 2021 — when nobody could afford to discover a problem a month late — made the case for it sharper than ever.
Why small batches win
The benefits compound, which is why this idea sits near the centre of Lean.
Shorter lead time. An item no longer waits for ninety-nine companions before it moves to the next step. Work flows through the process instead of sitting in a pile, so the first finished item arrives far sooner.
Faster feedback on defects. If a process produces an error, a batch of ten reveals it after ten units, not after a thousand. Small batches contain the blast radius of any mistake and shrink the rework.
Less work-in-process and less cash tied up. Large batches are inventory by another name — money and effort frozen mid-process. Smaller batches free that up and make the real state of the work visible.
Smoother demand on the system. Big batches arrive as spikes that overload the next step and then starve it. Smaller, more frequent handoffs level the load and reduce firefighting.
How to shrink batch size, step by step
Map the current flow and find where work accumulates — the queues in front of a step are usually batches in disguise.
Attack the setup cost that justifies the batch. The classic Lean move (SMED — single-minute exchange of die) is to make changeovers cheap enough that small batches stop hurting.
Cut the batch in half first, not to one. Halving is a safe experiment that reveals the real constraints before you over-commit.
Stabilize, observe, then halve again. Treat each reduction as a small test: measure lead time and quality, confirm nothing broke, then go smaller.
Standardize the new size once it holds, so the gain does not quietly drift back to the old comfortable pile.
Where it goes wrong
The usual failure is shrinking the batch without touching the setup cost. If a changeover takes two hours, smaller batches just multiply those two hours and everyone correctly concludes that small batches are 'inefficient.' Reduce the setup first, or in parallel, and the economics flip. The goal is not small batches for their own sake; it is fast, even flow with quick feedback — and smaller batches are the most direct route there.
If you want to find where batch size is quietly costing you time and cash, XNM's strategic advisory can help you map the flow and plan the change.