BOM Optimization Methods That Reduce Material Cost

Most of a product’s material cost is fixed while it is being designed, and the scope for reducing it later through purchasing is limited. Teams often focus on whether the function can be implemented and leave the cost of the bill of materials until production, where the margin turns out to be thinner than the business case assumed. BOM optimization is the discipline of deciding that cost earlier, and it works by choosing parts, not by arguing about prices.

Four Ways the Cost Comes Down

Material standardization means agreeing on a set of packages and parameter grades and using them across the product, which reduces the number of distinct part numbers and increases the volume bought against each of them. Fewer part numbers also means fewer feeder changes on the line, which is a saving at the assembly stage as well as at the purchasing stage.

Domestic substitution compares an imported device against a locally available equivalent on the parameters that matter, and selects the lower cost part where the comparison supports it. The saving on a controller or a connector can be substantial, and the comparison is made on data rather than on the origin of the part.

Supply chain work spreads the risk as well as the cost. A second approved source for a critical part, an agreement built on volume, and a review of the lead times all reduce the chance that the line stops for a part nobody can buy. Production fit is the fourth method: choosing packages and a layout that the line can place efficiently lowers the processing cost, and improving the board itself, in layer count or in material, lowers the cost of the bare board at the same time.

Working on the Cost at Every Stage

A target cost belongs at the start. The product has a cost that the business case requires, and that figure is divided among the functional blocks so that the controller, the power stage and the interfaces each carry a share. Without that division, the design is completed and then found to be over budget, and the correction is made under pressure at the worst possible moment.

During schematic design, the cost is decided part by part. Devices already in the standard library are preferred, a new part number has to justify itself, and the choice of a controller considers availability and the alternatives as well as the specification. Tolerance is one of the easiest places to find cost: a resistor that does not need to be within one percent can be a five percent part, and outside the circuits where the precision matters, the difference is invisible.

A design review with purchasing and quality present turns the bill of materials into a shared document rather than a list produced by one department. The hardware engineer speaks for the function, purchasing for the price and the availability, and quality for the reliability and the supplier. Suggestions from those three directions are collected before the list is frozen rather than after.

Prototyping is where substitution is proved. Where a part is intended to replace another, a comparison is arranged during the pilot so that the electrical behaviour, the temperature behaviour and the soldering behaviour can all be observed. A part is only entered into the production list once that comparison supports it, which is what keeps a cost decision from becoming a reliability decision.

Optimisation does not stop at volume production either. Reviewing the material every quarter shows which part numbers carry the most value, and a new device that offers a better price or a better supply position can trigger a change assessment. A bill of materials is a living document rather than a record of a decision taken once.

What the Methods Are Worth, and What They Cost

The methods differ in both the saving and the risk. Standardising packages and parameter grades typically removes five to ten percent with little risk, and the earlier it is done the more effective it is. Substituting a domestic part for an imported one can remove twenty to fifty percent, and it carries a moderate risk because the comparison has to be made carefully.

Reducing a specification, such as the tolerance of a resistor or the voltage rating of a capacitor, typically removes ten to thirty percent where the circuit genuinely does not need the higher grade. Changing the architecture of a block, such as moving to a different controller family, removes fifteen to thirty percent and carries a high risk, because it changes the design rather than the list.

Buying several products together against a single supplier removes five to fifteen percent with little risk. Choosing packages that the line can place efficiently removes five to ten percent with little risk as well. Reading the table as a whole is what makes the exercise useful: a method with a large saving and a high risk is not automatically the right one to choose.

The Limits That Are Not Crossed

Optimisation is not the same as looking for the cheapest part. A substitute has to meet the parameters the circuit requires, so the key specifications are not reduced. The temperature range, the life and the immunity to interference have to be at least equal to the original, because those are the properties that determine whether the product survives its application.

A part that is about to be discontinued or whose supply is unstable is not a saving, whatever its price today. Every substitute is verified before it is used, and every change is recorded with the reason for it and the review that approved it, so that the history of the design can be read afterwards.

Our component procurement group carries the approved list and the substitution records, and the design work is done by PCB design and layout with the resulting boards built through PCB manufacturing and turnkey assembly.

How to Run the Exercise

`r`n

The most useful habit is to treat the bill of materials as a deliverable rather than as a by-product of the schematic. It is issued with the design, reviewed by the three functions, and revised with a recorded reason. Each revision states which reference designators changed, what the change was, and what the expected effect on cost and on availability is, so that the history can be read a year later without anyone having to remember it.

`r`n

A second habit is to sort the list by value rather than to review it from the top. A handful of part numbers normally accounts for most of the material cost, and spending an afternoon on them produces more than spending a week on the long tail of small passives. The long tail is dealt with by standardisation, which is a single decision rather than a series of negotiations.

FAQ

Is optimisation just replacing parts with cheaper ones? No. Standardisation, buying together, specification review and process fit all reduce cost without changing the function, and substitution is only one of the methods.

Can a finished design still be optimised? Yes, but the scope shrinks with the stage. A schematic stage offers every method, a completed prototype offers substitution and specification review, and a product already in production offers buying and sourcing improvements.

How is the work charged? Where the project is a full development, the optimisation is part of the service. Where only the bill of materials is being reviewed, the work is quoted against its complexity and may be linked to the saving achieved.