Variant proliferation happens gradually. One customer wants a different voltage. Another wants their branding. A third needs a different connector. Each request is small and profitable, and five years later you have four hundred part numbers describing what is essentially eleven products.
The BOM problem this creates is real: maintaining four hundred near-identical structures means every common change has to be applied four hundred times, and it will be applied inconsistently.
The Configurable Alternative
Rather than one BOM per variant, you define a single structure with rules. The common components appear once. Variable positions carry conditions: if voltage is 240, use this transformer; if the market is European, use this plug set; if the customer is X, use this label.
A configuration — a set of chosen options — resolves the rules into a specific buildable BOM. Change a common component once and every variant inherits it.
Getting the Option Structure Right
This is where configurable BOMs succeed or fail, and it’s a modelling exercise rather than a systems one.
Options should be independent where possible. Voltage, colour, and connector type as separate dimensions is manageable. Bundling them into combined codes multiplies the option list and defeats the purpose.
Constraints between options need explicit rules. If the high-power motor requires the heavy-duty housing, encode that. Otherwise sales will configure combinations that can’t be built, and the error will be found at assembly.
Keep the rules readable. Deeply nested conditional logic becomes unmaintainable quickly — if only one person understands the configuration rules, you’ve moved the risk rather than removed it.
Planning Implications
Configurable products change how forecasting works. You forecast the base product and the option mix separately, which is usually more accurate than forecasting each variant, because option ratios are more stable than variant volumes.
It also enables postponement. Build or stock the common base, apply the differentiating options to order. This pools your demand variability across all variants and cuts the inventory needed for the same service level — typically the single largest benefit of the whole approach, and one that gets overlooked because it’s a supply chain gain rather than a data-management one.
When Not to Bother
Configurable BOMs carry setup and maintenance cost. If you have twelve variants that rarely change, discrete BOMs are simpler and perfectly adequate. The threshold is roughly where the number of variants times the frequency of common changes starts to exceed what anyone can maintain reliably.
They’re also poor fits for genuinely bespoke work where each order differs structurally rather than by option selection. Engineering-to-order needs a different approach.
The Variant Question Worth Asking
Separately from the BOM structure: how many of your variants earn their existence? Run volume and margin by variant. Most manufacturers find a long tail selling in single digits annually, each carrying inventory, tooling, documentation, and complexity cost. Rationalising the tail is often worth more than optimising how you manage it.
ticktick.ai supports rule-based configurable BOMs with option constraints, and plans base and option demand separately.
