Every shop has a routing template. Most of them are wrong.
Not wrong in a way that crashes the ERP. Wrong in the way that quietly bleeds margin: setup times that don’t match reality, missing operations nobody remembers to add, handoffs between departments that take twice as long as the template says. The quote looks clean. The job hits the floor. And suddenly you’re explaining to the customer why lead time slipped and margin evaporated.
The problem isn’t that shops don’t know their processes. It’s that routings get built once — often by someone who hasn’t run a machine in years — and then calcify. Meanwhile, the shop evolves: new machines, different tooling, changed staffing, updated quality requirements. The routing stays frozen.
What a routing actually needs to capture
A useful routing isn’t just a sequence of operations. It’s a map of how time, material, and information move through your specific shop. At minimum, it should reflect:
- Real setup and teardown — not the catalog number, but what your team actually spends on fixture loading, tool touch-off, first-article inspection, and changeover between jobs
- Queue and move times — the hours parts sit waiting for the next machine, the next operator, the next inspection
- Decision points — where an operator chooses a path (rework? scrap? continue?) and how that branches the routing
- Shared resources — the CMM, the heat-treat oven, the one programmer who knows the 5-axis CAM — and how contention for them stretches lead time
- Quality gates — FAI, in-process checks, final inspection — with realistic durations, not wishful thinking
If your routing doesn’t include these, your quote is a guess dressed up in math.
Start with a walk-through, not a spreadsheet
The best routings come from standing on the floor with a stopwatch and a notebook. Follow a job from receiving to shipping. Time every touch. Note every wait. Ask operators: “What happens if this tool breaks? What if the first article fails? Where does the job go next — and who decides?”
You’ll find things no spreadsheet captures:
- The 45 minutes the laser operator spends nesting parts because the nesting software doesn’t talk to the ERP
- The two-hour wait for the only inspector certified on AS9100 first articles
- The half-day the programmer spends fixing a STEP file the customer swore was “production-ready”
Document these. They’re not exceptions. They’re your process.
Build routings by process family, not by part number
Shops that maintain a unique routing per part drown in maintenance. Instead, group work into process families — “CNC mill 3-axis,” “laser cut + bend + powder coat,” “SLS print + bead blast + dye” — and build a master routing for each family.
Each master routing captures the standard sequence, typical setup bands, and common quality gates. Then use parametric overrides for part-specific variables: feature count, tolerance tier, material hardness, surface finish spec. This keeps the routing library manageable while still producing accurate estimates.
Solvi’s quoting engine https://www.solvi.io is built around this exact approach: process families with parametric overrides, so you define the logic once and apply it across thousands of part variations.
Capture the handoffs between departments
The biggest lead-time killers hide in the white space between operations. Programming to setup. Setup to first article. First article to production run. Production to inspection. Inspection to shipping.
Each handoff has a cost: communication delay, information loss, prioritization conflicts. A routing that treats these as zero-time transitions will always underestimate lead time.
Model handoffs explicitly. Add a “transfer” operation between departments with a realistic duration buffer. Track which handoffs consistently overrun — those are your improvement targets.
Version control your routings like code
When a new machine arrives, when a process changes, when you learn from a scrap event — update the routing. Immediately. With a change log.
Shops that treat routings as living documents catch drift early. Shops that treat them as “set and forget” discover the drift when a job loses money and someone finally compares actuals to estimate.
Link routing versions to quote versions. When you re-quote a repeat job, you’re using the current best knowledge, not the routing from three years ago.
Use actuals to close the loop
A routing is a hypothesis. Shop floor data is the test.
Capture actual setup times, cycle times, move times, and scrap rates per operation. Compare them to the routing. When variance exceeds a threshold — say 15% — flag the operation for review. Over time, your routings converge to reality.
This is where a connected MES pays off: it feeds actuals back into the quoting engine automatically, so the next quote inherits the lesson. No manual re-entry. No “I’ll update it later” that never happens.
Conclusion
Routings that match your shop floor aren’t a luxury. They’re the difference between quotes you can stand behind and quotes you hope nobody orders.
Start with one process family. Walk the floor. Time the reality. Build a master routing that reflects what actually happens — including the waits, the decisions, the handoffs. Then version it, measure against it, and improve it.
Solvi helps digital manufacturers build and maintain routings that stay accurate as the shop evolves — connecting quoting, MES, and capacity visibility in one platform. See how it works.