Build-to-print: making to someone else's drawing and owning none of the design

Conformance is the whole of the promise

You are committing to produce an article that matches a specification you had no hand in setting, at a revision level the customer controls and can change. There is no design latitude: a dimension that seems needlessly tight is still a requirement, and improving on it without approval is a nonconformance rather than a favour. Quality evidence is the deliverable alongside the part, which typically means first-article verification against every characteristic, material and process certification, and records that survive audit long after the parts have shipped. The customer retains design authority and the intellectual property, so the shop's asset is its demonstrated capability, not the product it makes.

Buying to a specification somebody else wrote

Material and special processes are usually constrained by the customer, sometimes to a named approved source, which removes most of the negotiating room a buyer would normally have. What remains is scheduling and documentation: getting certified material with the right traceability, on time, at a revision that matches the drawing. Outsourced treatments such as heat treatment, plating, painting and non-destructive testing sit on the critical path and are frequently approved by the end customer rather than by you, so switching one requires their agreement. Substituting anything to protect a date, without written concurrence, is the single fastest way to convert a delivery problem into a contractual one.

The work this attracts, and the measurement it obliges you to own

Demand comes from firms that hold the design and want capacity, and it clusters where documented conformity is contractually required: aerospace structures and details, defence, medical components, energy equipment, automotive tooling. The equipment profile is general purpose in production and unusually heavy in inspection, because proving conformance is as capital-intensive as producing it. Coordinate measurement, surface and thread verification, calibrated hand instruments under controlled recall, and often environmentally controlled inspection space all have to be funded from work that is priced per part. Shops that under-invest here win jobs on price and lose them again on rejected first articles.

How print work goes wrong, and what is left on the floor when it does

The classic failure is quoting from an incomplete pack: notes referencing specifications you have not read, a tolerance scheme that demands fixturing nobody costed, or a revision superseded between enquiry and order. Next comes silent interpretation, where an operator resolves an ambiguity sensibly and consistently, and the customer disagrees after delivery. Inventory exposure is real despite the model looking order-driven, because customer-specified material bought for a print job has no alternative use if the programme is cut. Shops protect themselves through material clauses tied to release schedules, and by refusing to hold stock against a forecast that carries no commitment.

Growth limited to capacity, and the document control that permits it

You scale by adding machines, people and inspection throughput, which is honest but linear, and price pressure is constant because the customer can move a drawing pack to another qualified shop. The defensible positions are qualification depth, on-time record and the ability to absorb a whole part family rather than a single number. Systems have to carry revision control at their core: which drawing revision produced which serial, which certificates apply, and what changed between releases. A shop that cannot answer those questions quickly will eventually ship to a superseded revision, and that single event costs more than the document system it declined to buy.

Frequently asked questions

Who carries the risk if the customer drawing itself is wrong?

Normally the design owner, but only if you flagged it. A shop that spots a conflicting dimension or an impossible tolerance and raises it in writing before manufacture has protected itself. A shop that notices and builds anyway, or notices nothing because it never checked the pack for internal consistency, ends up sharing the cost regardless of who drew the error. Treat drawing review as a chargeable, scheduled activity at order acceptance rather than something an operator does at the machine.

Can a build-to-print shop suggest design improvements?

Yes, and it is often the most valuable thing offered, but it belongs in a separate conversation from the build. Propose changes as a documented request with the manufacturing benefit stated, and manufacture to the existing revision until the customer issues a new one. The trap is informal agreement, where an engineer on the customer side says a change is fine and nothing is issued. When the parts are later measured against the released drawing, verbal permission has no standing.

What should be nailed down before quoting a print package?

The complete specification tree, including every referenced document and its revision, since those references frequently carry more cost than the drawing itself. Then the inspection requirement, because sampling and full-characteristic verification are different businesses. Then material source constraints and any customer-approved processors, which determine lead time. Finally, who owns tooling and fixtures and what happens to them if the programme moves. Quotes that skip these are estimates of the machining only, and the machining is rarely where the money goes.