Quality & Compliance
APQP: The Five Phases Before Your Part Ships
The short answer
Advanced Product Quality Planning is a structured five phase method for taking a part from concept to production without discovering the problems at launch. PPAP is not a separate exercise: it is the output of phase four, the evidence that the plan worked. Most buyers request PPAP without ever running an APQP, then are surprised when the submission arrives late and incomplete. The phases are planning, product design, process design, product and process validation, and feedback and corrective action.
The five phases
APQP came out of the automotive industry and is now the default framework wherever a part has to be launched into volume production with evidence behind it. The phases are sequential, and each one produces artefacts the next phase consumes.
The column that matters most to a buyer is the last one. Every phase has buyer obligations, and a supplier cannot complete a phase while waiting on you.
| Phase | Purpose | Key outputs | What the buyer owes |
|---|---|---|---|
| 1. Plan and define | Turn the customer requirement into a program | Design goals, reliability targets, preliminary bill of materials and process flow | Requirements, volumes, timing, special characteristics |
| 2. Product design and development | Finish a design that can be made | Design FMEA, drawings, material specifications, DFM review | Released drawings, tolerance rationale, critical dimensions identified |
| 3. Process design and development | Design the process that will make it | Process flow, process FMEA, control plan, packaging standard, work instructions | Approval of packaging, labelling and any customer specific process requirement |
| 4. Product and process validation | Prove it works at rate | Significant production run, MSA, capability studies, PPAP submission | Prompt PPAP review and disposition, gauge approval |
| 5. Feedback, assessment, corrective action | Improve what production reveals | Reduced variation, closed corrective actions, updated control plan | Timely defect and field data back to the supplier |
Where PPAP actually sits
PPAP is the deliverable at the end of phase four. It is the package of evidence proving that the process designed in phase three, run on production tooling at production rate, produces parts that meet the drawing from phase two.
This is why a PPAP request that arrives with no prior planning goes badly. The submission needs a design FMEA, a process FMEA, a control plan, measurement system analysis and capability data. Those are phase two and three artefacts. If nobody produced them along the way, the supplier is reconstructing three phases of work retroactively while you wait, and the reconstruction is documentation rather than the thinking the documents were meant to capture.
If you want a real PPAP, ask for the phase three artefacts before the production run, not after.
What the buyer owes, and when
The most common cause of a late launch is not supplier performance. It is the buyer treating APQP as something the supplier does alone.
In phase one the supplier needs volumes and timing, and they need to be honest numbers. A supplier who plans for 500 parts a year and receives an order for 20,000 has tooled the wrong process. In phase two they need released drawings with critical dimensions identified and, ideally, the reasoning behind the tight ones, because a tolerance with a stated purpose can be negotiated and one without a purpose cannot.
In phase three they need decisions from you: packaging, labelling, any process you require or prohibit. In phase four they need you to actually review the PPAP. A submission that sits unread for six weeks stalls the launch just as effectively as one that was never sent, and it is a genuinely common failure.
In phase five they need your field data. A supplier who never learns which parts failed in service cannot reduce the variation that caused it.
The control plan outlives the launch
Of everything APQP produces, the control plan is the document with the longest useful life. It lists each process step, the characteristic being controlled, the specification, the method of measurement, the sample size and frequency, and the reaction plan when something goes out of limits.
It is the document that answers the question you will actually ask in eighteen months, which is why did this lot come out different. If the control plan says the bore is checked with an air gauge every twenty parts and the reaction plan is to quarantine back to the last good check, you can reconstruct what happened. Without one you are guessing.
Ask for the control plan even when you are not running a full APQP. It is one page, every competent shop can produce one, and it is the highest value document per unit of effort in the entire framework.
APQP scaled down for low volume
The full framework was designed for automotive programs producing hundreds of thousands of parts, where a launch failure costs millions. Applying it unmodified to a 500 unit hardware program is theatre, and suppliers will price it accordingly.
The scaled version keeps the thinking and drops the paperwork. Identify your critical characteristics, the handful of dimensions and properties that actually determine whether the part works, and say so on the drawing. Ask for a process flow and a control plan covering those characteristics only. Run a first article inspection instead of a full capability study. Ask for a process FMEA only on the operations that produce the critical characteristics.
That is perhaps a tenth of the effort and captures most of the protection, because the protection was always concentrated in a few features.
What this costs you in time
APQP timelines are driven by tooling and validation, not by paperwork, and the paperwork is often blamed for delays that came from a late design freeze.
The durations below are typical for a machined or molded part at moderate volume. The one that surprises people is phase four, because the production run has to happen on production tooling, and tooling lead time is the gate.
| Phase | Typical duration | Usual cause of delay |
|---|---|---|
| 1. Plan and define | 1 to 3 weeks | Volumes and timing not settled by the buyer |
| 2. Product design | 2 to 8 weeks | Design changes after the supposed freeze |
| 3. Process design | 2 to 4 weeks | Waiting on packaging and process decisions from the buyer |
| 4. Validation | 4 to 16 weeks | Tooling lead time, then buyer review of the PPAP |
| 5. Feedback | Ongoing | No field data returned to the supplier |
When APQP is overkill
Skip it for prototypes. Skip it for parts you will order once. Skip it where the part is not safety relevant, the volume is low, and a first article inspection plus a dimensional report gives you the evidence you actually need.
Run it where a launch failure is expensive: hard tooling you cannot cheaply modify, a regulated product, a customer who will audit your supply chain, or a volume high enough that a process shifting out of center produces thousands of bad parts before anyone notices.
The honest test is what happens if the first production lot is wrong. If the answer is that you scrap 200 parts and reorder, a first article is proportionate. If the answer is that you scrap a tool, miss a launch window, or file a regulatory deviation, plan the launch properly.
What to send, and what to ask
Send released drawings with critical characteristics marked, your real annual volume and its expected ramp, your timing including the date you actually need parts, and any customer specific requirement you are passing down.
Then ask for four things. A process flow diagram, so you can see how the part is made. A process FMEA covering the operations that produce your critical characteristics. A control plan naming what gets measured, how often, and what happens when it drifts. And a clear statement of which PPAP level you will receive and when.
If a supplier can produce those four before the production run, the PPAP at the end is a formality. If they cannot, you have learned something important early, which is the entire point of planning.