Production Phase in Product Development

Production is the phase where the product is made in quantity and sold. The question it answers is can this be made and shipped at volume, reliably? It is the last phase, and the only one that does not end. As long as the product is on sale, it is in production, and it keeps needing engineering.

What changes at production

Scale. Processes that worked for a pilot run of fifty may not hold at five thousand a month. Supplier capacity, assembly line balance, incoming inspection and packaging all get stressed in ways the pilot did not reveal.

Where this phase sits. The highlighted marker is this phase. Everything to its left has been answered; everything to its right is still a range.

Consistency. The goal shifts from "does it work" to "does every one work, every time." Quality control becomes a system: inspection points, sampling plans, test fixtures, records, and a defined response when something fails.

Supply chain. Parts come from many suppliers with their own lead times, minimum orders and occasional surprises. A single component going end-of-life can stop a product line. Someone has to watch for this.

Cost pressure. The measured unit cost from Preproduction becomes the baseline, and the ongoing work is to lower it: better yields, negotiated prices, design simplifications learned from the line.

Field feedback. Real users find things that testing did not. Returns, reviews and support calls become engineering inputs.

Ongoing engineering

Production engineering is a different kind of work from development, but it is still engineering:

Sustaining engineering. Fixing problems found in the field, replacing obsolete components, qualifying second sources for critical parts.
Cost reduction. Revisiting the design with volume data in hand. A part change that saves a dollar is worth ten thousand dollars a month at ten thousand units.
Variants and next versions. A new color, a new market with different certifications, a version 2 with the features that were cut from version 1.
Engineering change control. Every change to a part, process or document is proposed, evaluated, approved and recorded, so that the product being shipped matches the product that was tested.

Scoped work like this is often a good fit for a fixed price, because the product is known and the change is bounded.

Key takeaway The first shipment is not the finish line. It is the point where the product starts generating real data, and the data drives the next round of engineering.

What the owner needs in place

Documentation control. One current, controlled set of drawings, bill of materials and procedures. The factory builds what the documents say.
A quality plan. What gets inspected, how often, against what standard, and what happens on a failure.
Supplier agreements. Lead times, pricing tiers, quality expectations, and what happens when a part changes.
A support path. How customers report problems and how those reports reach engineering.
Ownership of everything. Design files, tooling, certifications, and the documentation. A product that cannot be moved to another factory is hostage to its current one.

Typical cost and duration

Development labor drops sharply once the product is stable, and the ongoing engineering spend becomes a fraction of unit sales. The dominant costs are now the product itself: parts, assembly, freight, duties, warranty. Those are business costs, not development costs, and they belong in a different part of the plan.

Common mistake Letting the engineering relationship lapse at the first shipment. The first component going obsolete, the first unexpected field failure, or the first big retailer asking for a change finds nobody who knows the design.

What to ask a manufacturer about this phase

"How are engineering changes proposed and approved?"
"What is the inspection plan, and what are the acceptance criteria?"
"How much notice will I get if a component is discontinued?"
"If I need to move production, what do I take with me, and what stays?"

Try this (15 minutes) Imagine the product has been on sale for a year. List three things that are likely to have gone wrong (a part change, a failure mode, a competitor move). For each, write who would handle it. Any blank is a gap in the plan.