Machinery manufacturing: selling a promise, then financing it through milestones

Engineering hours are the raw material

In this sector the scarce input is not steel but the design office. A specification that arrives half-defined consumes engineering capacity long before the shop sees a drawing, and every hour spent clarifying scope with a customer is an hour unavailable to the next order. Builders who track engineering hours per order against the estimate discover quickly which sales engineers are quoting fantasy. The discipline that matters is a scope freeze: a defined moment after which changes are quoted as variations rather than absorbed. Firms without one do not usually fail on manufacturing cost; they fail because their design office is permanently rebuilding orders that were sold as standard.

A job shop with one assembly bay at the end

The machining and fabrication side behaves like any high-mix job shop: setup dominates, scheduling is a daily argument, and the queue in front of the large machining centres decides the delivery date. What differs is the convergence point. Everything has to arrive at the assembly bay in the right order, because a missing bracket stops a build that occupies floor space no other order can use. Kitting before assembly starts, rather than issuing parts as they are called, is the single change most builders make after their first badly overrun contract. It exposes shortages while there is still time to expedite them.

Commissioning is where the estimate goes to die

Factory acceptance in a controlled bay is not the same as running on the customer's material, in the customer's building, with the customer's upstream equipment feeding it. Product variation, foundation quality, compressed air quality, operator training and integration with existing controls each add days. Site labour is expensive and travel time is unrecoverable. Experienced builders quote commissioning as a defined scope with clear customer obligations for site readiness, and they instrument the machine so remote diagnosis is possible before someone gets on a plane. Those who treat commissioning as a small percentage added at the end lose most of their project margin there.

The safety file is a deliverable, not paperwork

A machine placed on the market carries a conformity declaration from its builder, backed by risk assessment, protective measures and a technical file that has to exist before shipment rather than being assembled afterwards. Where a customer combines several machines into a line, responsibility for the assembly as a whole has to be assigned in the contract, and it is routinely left ambiguous. Guarding decisions made late are expensive because they affect access, ergonomics and cycle time. Builders who bring safety engineering into concept design ship on time; those who bolt it on after the machine runs end up redesigning access panels in the assembly bay.

Parts and service are the business that survives the cycle

New machine orders vanish in a downturn. The installed base does not. Wear parts, retrofits, control upgrades, calibration and breakdown response generate revenue that is less cyclical, higher margin and largely protected from price competition because the customer needs the machine running today. This requires deliberate work: serial number records, as-built configuration control, parts identification the customer can use, and a service organisation that is staffed before it is needed. Builders who never invested in that side discover in a weak year that their competitors are quietly servicing machines the original builder sold.

Frequently asked questions

Why does the first machine of a new design almost always lose money?

Because the first build is where design assumptions meet reality. Parts do not fit, sequences have to be reworked, controls need tuning that was not scheduled, and the assembly team is learning the machine while building it. None of that appears in an estimate built from a parts list. Sensible builders treat the first unit as development cost, recover it across the expected series, and only quote series pricing once a build has actually been completed and its hours recorded.

Should a machinery builder machine parts in-house or buy them?

In-house machining buys schedule control and protects design confidentiality on the parts that define machine performance. It also loads fixed cost into a business with irregular order intake, which is precisely the wrong shape. Most builders end up keeping the precision and prototype work in-house, subcontracting bulk and simple parts, and maintaining enough internal capability to make a supplier replaceable. The parts to keep are those where late changes are frequent, not those with the highest purchase value.

How do you price a machine that has never been built before?

By decomposing it into elements you have built and pricing the unknown parts as explicit risk. Bought-in content can be quoted. Fabrication and machining can be estimated from similar work. Engineering, assembly and commissioning hours are where estimates fail, so they should be built from historical actuals on comparable projects rather than from optimism. Adding contingency to the whole price hides the problem; identifying the genuinely novel elements and pricing those separately makes the risk visible to both sides.