Most manufacturers settle into one mode and apply it broadly. Make-to-order businesses make everything
to order; make-to-stock businesses stock everything. The strategy was chosen once, usually early, and it
persists because revisiting it feels like a strategic question rather than an operational one.
It’s actually a per-product decision, and it changes over a product’s life.

What Each Mode Costs You

Make-to-stock buys availability with inventory. You hold finished goods, so you ship immediately, and you
carry the cost and the obsolescence risk of guessing wrong about what customers will want.
Make-to-order buys capital efficiency with lead time. Nothing is built speculatively, so nothing is stranded,
and the customer waits. It also concentrates your capacity risk — demand spikes have to be absorbed by
the plant rather than by a buffer.
Neither is better. The right choice depends on three things: how predictable demand is, how long
production takes relative to what customers will accept, and how much variety you offer.

The Deciding Factors

Demand predictability. Stable, forecastable demand makes stock-holding low-risk. Erratic demand makes it
a gamble, and the gamble is usually lost on the tail items.
The lead time gap. If customers accept four weeks and you can build in three, make-to-order works with no
service penalty. If they expect delivery in two days and production takes three weeks, you must hold
something — the only question is what form it takes.
Variety. Stocking every variant of a wide range is expensive and produces dead stock in the tail. Stocking a
few high-volume lines is efficient.
Margin and obsolescence. High-margin products tolerate stock-holding cost more comfortably. Products at
risk of design change or expiry tolerate it badly.

The Middle Options

The most useful answers are usually between the two. Assemble-to-order holds sub-assemblies and
finishes on receipt of the order — the differentiation happens late, so you carry inventory that serves many
variants rather than one. This is the postponement idea, and where it’s technically feasible it usually beats
both pure modes.
Configure-to-order works the same way for products differing by option rather than structure. And a hybrid
across the range — top sellers to stock, tail to order — is often the right answer even when it feels
inconsistent.

Where Analysis Helps

The decision needs demand variability, forecast accuracy, margin, lead time, and carrying cost per item.
That’s a calculation nobody wants to run manually across a full catalogue, which is why the policy tends to
be set at category level and left.
Run per item and it will reclassify products you’d have assumed were settled. It should also be re-run
periodically, because a product’s right mode changes as it moves from launch through maturity to decline
— and the transition to make-to-order in decline is the step that prevents most end-of-life write-offs.
Expect resistance to a mixed policy. A single mode is simpler to explain and simpler to operate, and moving
to per-item classification means planners and sales both handling products differently depending on the
item. That operational cost is real and should be weighed — but it’s usually smaller than the inventory it
saves.
ticktick.ai evaluates the fulfilment mode per item against demand variability, lead time, and carrying cost,
and flags products whose position has shifted.