Guides / Process
Taking a Bike Part From One Piece to Five Hundred
September 18, 2026
Most machined bike parts follow the same path. The brands that get through it quickly are not the ones with the best CAD. They are the ones that know what each stage is for and refuse to skip a step.
Stage 1: Fit sample, 1 to 3 pieces
Purpose: confirm the part fits the bike and the assembly works.
Machine it as machined, no finish. Skip the cosmetic pockets if they slow the job. You want geometry in your hands this week, not a beautiful part next month.
What you learn here is almost always dimensional: a bolt that fouls, a clearance you forgot, a cable that routes differently in reality than in CAD.
What to lock after this stage: overall geometry and every interface dimension.
Stage 2: Function sample, 5 to 20 pieces
Purpose: ride it and test it.
Now machine it properly, in the production material and the production finish. Testing an unanodized part tells you about a part you will never sell, because anodizing reduces fatigue life.
This is also where fatigue testing happens if the part is structural. Testing consumes parts. Plan for that in the quantity.
What to lock after this stage: material, temper, finish type and thickness, and the design itself. This is the design freeze, and it should be an actual event with a date and a drawing revision, not a gradual fading of change requests.
Stage 3: Pilot batch, 25 to 100 pieces
Purpose: prove the process rather than the part.
Different question from stage 2. You already know the design works. Now you want to know whether the shop can make it repeatably, whether the anodizer can hold the color and whether the inspection routine catches problems.
What to do here:
- Ask for a first article inspection report with actual measured values, not a pass or fail sheet
- Approve a physical anodize master sample and keep one yourself
- Agree what happens to parts that fall outside tolerance
- Time the process honestly. Add up machining, finishing and transit, and use that number for planning, not the optimistic one
What to lock after this stage: the inspection plan, the anodize master, the packaging and the real lead time.
Stage 4: Production, 100 pieces and up
Purpose: repeat what worked.
By now nothing should be new. The things that go wrong at this stage are almost always changes somebody thought were minor:
- A different billet supplier, which changes anodize color
- A different anodizer, which changes everything about the finish
- A small design tweak, which invalidates the fatigue test
- A new operator following an incomplete drawing
This is why the drawing matters more than the relationship. A complete drawing, with tolerances, finish spec and inspection requirements, is what survives a change of shop, operator or season.
The two ways brands lose a season
Skipping the pilot batch. Going from 10 samples to 500 production pieces means the first time the process runs at volume is also the first time it runs for real. Color drift, a marginal fit or a deburring shortfall then applies to the whole order.
Changing the design after testing. A small change to a pocket depth feels safe. If the part is structural and you have already tested it, it is not safe, it is untested. Either accept the change and retest, or hold it for the next model year.
What to ask for at each RFQ
Say which stage you are at. A shop quoting a fit sample and a shop quoting a pilot batch are solving different problems, and a shop that knows 500 pieces are coming behind a 20 piece order will treat the fixture and the program as an investment rather than a cost to recover immediately.
Put the stage, the quantity and the realistic follow-on volume on the RFQ. It is the cheapest thing you can do to get a better price.
Frequently asked questions
Should the prototype and the production part come from the same shop?
It helps, because the fixture, the program and the knowledge stay in one place. It is not essential. What matters is that the drawing is complete enough that a second shop makes the same part, and if it is not, you were relying on the first shop's memory.
When should the design be frozen?
Before fatigue testing, not after. Every change after a successful test either invalidates the test or forces you to argue that it did not, which is a conversation nobody wins.
How many prototypes are enough?
Enough to fit check, enough to ride and enough to destroy. Fatigue testing consumes samples. Budget for the ones you will break rather than being surprised when the lab asks for five.
Need a quote for this part?
Send the drawing. We match you with up to 3 shops that machine bike parts and reply within 24 business hours.
Get matched with a bike parts shop