By the time your product reaches the production line, 80% of its cost is already decided. And almost nobody is talking about that 80%.

Here is the uncomfortable truth most cost review meetings never reach: you are almost certainly fighting over the wrong number.

By the time your product hits the factory floor, roughly 80% of its cost has already been decided—in the design room, in early material choices, in tolerance callouts, in requirements that nobody questioned. Manufacturing can only touch the remaining 20%. And yet that’s where almost all the energy goes.

Push suppliers harder. Move factories to lower-cost regions. Swap materials, downsize teams, chase waste on the production line. These efforts are real, but they are fighting over a fraction of the problem. The real leverage is upstream, in R&D—and most organisations leave it almost entirely untouched.

The 20:80 Rule: Understand It First

R&D decisions determine 80% of a product’s cost. Manufacturing decisions determine 20%. This is the foundation of design for cost—and until an organisation genuinely internalises it, every other initiative is treating symptoms.

Cost does not appear at the factory. It is designed in, layer by layer: through requirements, through the architecture chosen to meet them, through the parts selected, the materials specified, the tolerances written, the assembly sequence assumed. Remove an unnecessary requirement early, and you remove the entire cost tree that grew from it. Catch it late, and no amount of supplier negotiation will recover that ground.

Cost down is an R&D discipline. Not a procurement discipline. Not a manufacturing discipline. The earlier it happens, the more of the problem it can actually solve.

Start from Where Things Begin, Not from Where You Are

The most powerful shift in cost-down thinking is this: stop optimising the existing solution, and start from scratch with only the requirements in front of you.

When you start from where you are, you inherit everything—the assumptions, the workarounds, the accumulated over-engineering of every decision that came before. When you start from first principles, you ask a different question: what does this product actually need to do?

Tesla’s battery development is the example I keep coming back to. Rather than improving on an existing pack architecture, the team questioned every component from the physics upward. The result wasn’t an incremental saving—it was a fundamentally different structure. The same logic, applied to vehicle manufacturing, replaced a complex assembly of hundreds of stamped parts with a single cast unibody. Not because anyone found a cheaper stamping supplier. Because someone asked whether stamping needed to happen at all.

The chain to understand and act on is: requirement → design → manufacturing solution → material. Every link carries cost. Question each one.

Break the Walls in Your Head First

Before any methodology, any framework, any structured workshop—there are four mental walls that will kill your cost-down effort quietly and completely if you don’t address them.

Inertia — We’ve always done it this way. Don’t change what works.

Self-defence — Protecting past decisions, past designs, past territory.

Short-sightedness — Solving the immediate constraint, missing the system.

Passivity — Waiting for someone else to raise the issue.

Senior engineers are often the most valuable people in the room—and also the most limited by inertia. The phrases “we always do it this way” and “our competitors do it this way” are habits dressed up as engineering arguments. They are not reasons.

Elon Musk’s framework cuts through this cleanly: Question. Delete. Simplify. Accelerate. Automate. In that order. Most teams skip straight to optimising something—when the right first move was to delete it entirely. A requirement that doesn’t exist can’t cost you anything.

So ask the uncomfortable questions. Why does this part need to be this thick? Why this material? Does this feature actually appear in the product requirements, or did it materialise in a meeting? Can the part be thinner, smaller, combined with the one next to it? Can the connection method change—from screws to welding, from welding to a snap fit?

And watch for over-engineering. A product rated to IPX4 that was designed to IPX7 carries unnecessary cost in every single place that decision touches. The same goes for tolerances tighter than the function demands, and for jigs designed to inspect variation that should have been designed out of the process entirely.

Look Sideways—The Answer Often Already Exists

One of the most underused moves in cost reduction is looking at other industries. Most problems in product design have been solved somewhere already—just not in your category. The solution that saves you six months of development and a significant chunk of BOM cost might be sitting in medical devices, aerospace, or consumer electronics, waiting to be borrowed.

This is why the first step in any serious cost-down effort should be proper research. Map the full solution space before you start generating ideas: what materials exist, what manufacturing techniques are available, what functional alternatives have been tried, what benchmark products look like from the inside.

A Real Process, Not Just a Workshop

Brainstorming sessions are fine for warming up a room. They’re not a methodology. Cost reduction at the R&D stage needs a systematic process—one that ensures nothing gets missed and good ideas actually make it through to implementation.

  1. Research first. Before any ideas, map the full solution space—materials, techniques, functional alternatives, benchmarks across your own category and adjacent ones.

  2. Break the product down to parts level. For each part, look at it from three angles: what it actually does, why that material was chosen, and what the manufacturing process requires.

  3. Get cross-functional input. NPI, suppliers, service, quality, procurement—each sees cost differently. Suppliers especially will often have solutions they can only offer if you ask. Don’t just squeeze them. Help them find the answer with you.

  4. Break down the cost structure explicitly. Material, machine time, tooling amortisation, process yield, assembly labour. List it out. You cannot reduce what you haven’t made visible.

  5. Verify solutions properly. Use FEA for structural optimisation—it’s one of the most powerful cost-down tools available and still underused. Use 3D printing to test alternatives before committing to tooling. Rate each proposal by difficulty and impact, then prioritise ruthlessly.

One More Thing Worth Saying Out Loud

Some engineers design things expensively—and then cost them down later, in a way that reflects well on them personally. It’s a career move masquerading as engineering. The best defence is a culture where cost visibility is built into every design review from the start, not examined under pressure when margins have already shrunk.

And finally: keep learning. New materials, new manufacturing processes, new simulation tools arrive constantly. The answer to a cost problem you’ve been carrying for years might have arrived last quarter—in a different industry, in a research paper, in a startup you haven’t heard of yet.

Design for cost isn’t a project you run once. It’s a discipline—one that takes the right questions, the right process, and the willingness to challenge decisions that everyone stopped questioning a long time ago. The companies that get this right aren’t squeezing suppliers. They’re designing cost out before it ever gets built in.

Over to you: What’s the most expensive design decision you’ve seen that nobody questioned until it was too late—and what finally broke the wall?