By the time a consumer electronics product is in production and someone asks "can we bring the BOM cost down," most of the leverage to do so cheaply is already gone. Component footprints are locked into the PCB layout, firmware has been written against specific part behavior, and any change now means requalifying a design rather than making a choice. BOM cost is much easier to manage as a design constraint from the first prototype than as a cost-reduction project applied afterward.
Component consolidation is cheap early and expensive late
Every unique part number on a BOM carries cost beyond its unit price: it's a separate line in the pick-and-place program, a separate inventory SKU, a separate qualification and testing burden, and a separate point of supply-chain risk. Early in a design, it's common to accumulate parts opportunistically — a slightly different resistor value here, a different connector variant there — because each individual choice seemed locally reasonable. Consolidating to a smaller set of standard, well-stocked parts (common passive values, a single connector family, a single MCU family across a product line where possible) reduces both unit cost through better volume pricing on fewer parts and the hidden operational cost of managing more SKUs.
Doing this consolidation pass during schematic capture costs almost nothing. Doing it after the board is laid out, tested, and in production means re-qualifying every changed part and re-running whatever testing validated the original design — a cost that easily outweighs the unit-price savings for anything but a high-volume product.
Second-sourcing: a resilience decision that looks like a cost decision
Second-sourcing — qualifying a pin- or function-compatible alternate part from a different vendor — is often framed purely as a cost lever, but its real value is usually supply resilience. A single-sourced part, especially a specialized IC, can become an availability bottleneck during a component shortage or a vendor's own supply disruption, at which point production stalls regardless of unit cost. Building second-source qualification into the design and procurement process from the start — even if the alternate part is never actually purchased — gives a production line an escape hatch that's very expensive to build after a shortage has already stopped the line.
The tradeoff is that second-sourcing takes engineering time to qualify and sometimes requires design margin to accommodate two parts' worth of electrical tolerances instead of one. For high-volume consumer products where a supply interruption is costly, that time is usually worth it; for very low-volume or short-lifecycle products, it may not be.
The cost curve from prototype to 10k to 100k units
Unit costs don't scale linearly with volume, and the components that dominate cost at each stage change. At prototype quantities, NRE (non-recurring engineering) costs like tooling and one-off part purchases dominate, and component choice is driven mostly by availability and development convenience. At low-volume production (order of thousands of units), component pricing starts to matter but minimum order quantities and per-reel pricing breaks can distort the picture — a part that looks cheap at 100 units can be a poor choice at 5,000 if it doesn't hit a favorable pricing tier. At high volume (tens or hundreds of thousands of units), component selection, tooling amortization, and yield all compound, and small per-unit savings become significant in aggregate — which is also when custom tooling (injection-molded enclosures, custom connectors) starts to make economic sense in a way it didn't at lower volumes.
Designing with a target production volume in mind — even an approximate one — helps avoid choosing parts that are optimal at the wrong point on that curve.
Practical takeaway: cost optimization has to run alongside DFM, not after it
Design for Manufacturing (DFM) and BOM cost optimization are often treated as sequential steps — design the product, then optimize cost, then optimize for manufacturing. In practice these three pull on the same levers: a component choice that improves assembly yield often also reduces cost through consolidation, and a layout choice that eases manufacturing often also simplifies the BOM. Running them together, starting from the first prototype, produces a design that's cheaper to build and easier to scale than retrofitting cost reduction onto a design that was optimized for neither. See our embedded hardware work for how we approach this during design.