The project that starts when sales are working
In first-generation development, the priority is getting to market — correctly so. The consequence is that almost every product reaches production carrying fat: components chosen because they were available, tolerances tightened for safety, and solutions that were right for a prototype and simply never revisited. Once sales stabilize, that fat becomes a gold mine. Every dollar removed from unit cost is a dollar of margin, multiplied by every unit you will ever sell, forever. The systematic way to go after it is called value engineering.
The principle: value, not just cost
The defining rule is that you reduce cost only where the customer cannot feel it. Value engineering asks two questions about every single component: what function does this deliver for the customer, and what does it cost? An expensive part delivering a marginal function is a prime candidate. A cheap part that touches what customers love about the product is off limits. A cost reduction the customer experiences as declining quality is not a saving — it is brand damage that shows up later as returns and reviews.
Where to start: Pareto analysis of the BOM
Break the product down by cost. Almost invariably, five to ten line items account for the majority of the total. That is where you concentrate: a ten percent improvement on the largest item is worth more than eliminating twenty trivial ones. Build the structure properly first — see how to build a bill of materials — and make sure the cost model includes freight, duty, scrap, and assembly labor, not just purchase price. On imported goods, duty and freight can shift the priority order substantially, so work from landed cost rather than ex-works price.
The toolbox
- Part consolidation. Two parts that are assembled together and never move relative to each other are candidates to become one. Every part eliminated removes a purchase order, an inspection, an assembly step, and a failure mode.
- Tolerance relief. Every tight dimension that is not functionally required raises machining cost and scrap rate. Going through the drawings and asking “why is this tolerance here?” is often the cheapest win available — see GD&T basics for how to specify what actually matters instead of tightening everything.
- Material or process substitution. A machined part that becomes molded, a metal part that becomes engineering plastic, a fabricated assembly that becomes a single casting. At the right volume the difference is dramatic — the economics of tooling amortization are laid out in injection molding costs.
- Off-the-shelf instead of custom — and sometimes the reverse. At low volume a purchased module beats a custom design. Past a certain volume, a purpose-built part is cheaper than the module you were buying. Both directions are legitimate; the volume tells you which.
- Renegotiation and smarter sourcing. If actual volumes are running above the forecast the original pricing was based on, go back to your suppliers. Consolidating spend, committing to schedules, and qualifying a second source all move price.
- Assembly and test cost reduction. A better fixture, a self-locating feature that removes an alignment step, a test sequence that runs in half the time. Labor and test minutes are real BOM lines even when they never appear on the BOM.
- Design for manufacturing, applied retroactively. Most of the classic techniques are in design for manufacturing; a second pass over an existing design usually finds plenty that was missed under launch pressure.
Three traps to avoid
- Damaging reliability. A cheaper component that raises field failure rates erases the saving through warranty costs and reputation. Every change goes through durability and reliability testing before it ships — the qualification burden is part of the cost of the change, and it should be in your business case.
- Triggering re-certification. Swapping a power supply, a radio module, or a flame-rated plastic can invalidate existing test reports and force re-testing that costs more than the annual saving. Check this before you commit — see product safety testing requirements.
- Undocumented change creep. A series of individually harmless “minor” changes, none of them recorded, produces a product nobody can reproduce and failures nobody can trace. Every change — including the obvious ones — goes through a controlled process with a revision, a reason, and a validation record.
Suppliers are partners, not adversaries
The richest source of cost reduction ideas sits with your suppliers, and they usually will not volunteer it. The injection molder knows which small geometry change would cut cycle time by fifteen percent. The assembly house knows which connector saves a minute of labor per unit. The component distributor knows about a newer, cheaper equivalent that was not available when you designed.
Two mechanisms extract that knowledge:
- A joint cost-reduction workshop. Sit down with the supplier, put the product on the table, and ask directly: what would you change to make this cheaper to produce?
- A shared-savings incentive. Agree in advance to split the savings from any idea the supplier brings. This flips their incentive from staying quiet to actively hunting, and it converts an order-taker into a thinking partner. It is one of the quiet assets of a long manufacturing relationship.
Verify that the resulting parts still meet spec with the same rigor as a new product; the sampling framework is in AQL inspection explained.
The metric: a few percent per year
In mature industries, an annual unit-cost reduction of a few percent is a built-in expectation, and it is what allows a company to defend margin as market prices erode. A company that treats unit cost as a fixed input discovers a few years later that competitors sold at the same price and earned more. Set a target, own it, and review it on a schedule.
Remember that cost and price are separate decisions. Reducing cost gives you options — hold price and take margin, cut price and take share, or reinvest in features — and which option is right depends on your channel math. That reasoning is laid out in how to price a product.
Run a cost-down pass on your product
Projects House runs value engineering programs on products already in production — BOM teardown, redesign of the expensive items, supplier workshops, and the qualification testing that makes the changes safe to ship. If your product is selling and its margin is thinner than it should be, send us the details through our contact form. Related material is in our new product development section.