← Back to Field Notes

September 8, 2025 · The Bloomfield Team

A Step-by-Step Guide to Mapping Manufacturing Workflows

Most manufacturing leaders can describe their process at a high level: we get an RFQ, quote it, win the job, schedule it, run it, ship it. That description covers roughly 10% of what actually happens between receiving a customer inquiry and loading a box on a truck. The other 90% is where the time, money, and quality problems live.

Workflow mapping is the discipline of documenting every step, every handoff, every decision point, and every system involved in moving a job through your operation. Done well, it reveals bottlenecks that are invisible at the high-level view.

Step 1: Pick One Process and Follow One Job

The mistake most shops make is trying to map everything at once. Start with a single process. Quoting is the most common starting point because it touches the most systems and the impact of improving it shows up fastest in revenue.

Pick a real job that went through your shop recently. A medium-complexity part for a repeat customer. Trace that specific job from the moment the RFQ arrived to the moment the quote went out the door. Every person who touched it, every system they used, every piece of information they needed and where they found it.

For a broader view of how process mapping connects to operational improvement, see our guide to production visibility.

Step 2: Document the Information Flow

The physical flow of a part through the shop is usually well understood. Raw material arrives, goes to first op, second op, inspection, finishing, packaging, shipping. Operations teams have been mapping that sequence for decades.

The information flow is the part most shops have never mapped. Where does the drawing come from? What format is it in? Who reviews it first? Where does the material specification get confirmed? What system holds the customer's purchase order terms? Where does the schedule live? Who decides priority when two jobs compete for the same machine?

Build a simple table with four columns: Step, Person, System/Tool, Time. Walk through the process and fill in each row. The system column is where the mapping gets interesting. You will find jobs where information passes through seven or eight different tools: ERP, email, spreadsheet, phone call, paper form, whiteboard, walk to the floor, verbal confirmation.

Step 3: Mark the Wait States

In most manufacturing workflows, the actual work takes a fraction of the elapsed time. Quoting a part might involve 90 minutes of focused estimating work. The elapsed time from RFQ receipt to quote delivery is three days. The gap is wait time.

An estimator is waiting for material pricing. The quote is waiting for approval. A scheduler is waiting for confirmation that a machine is available. A quality record is waiting for the inspector to return from lunch and enter the data someone measured two hours ago.

Mark every point where a job stops moving and the reason it stops. These wait states are the highest-leverage improvement targets because eliminating them compresses cycle time without changing how the actual work gets done. The origin of most late deliveries traces back to information wait states in the front office, not capacity constraints on the floor.

Step 4: Identify the Decision Points

Every workflow contains branch points where someone makes a call. Can we hold the tolerance on this feature? Should we quote this as a three-op or four-op job? Does this job need first-article inspection? Is the delivery date achievable with current backlog?

Map each decision point and record who makes it, what information they need to make it, and where that information comes from. The goal is to determine which decisions require human judgment (estimating a fair price on a complex part) and which decisions could be automated or pre-populated (looking up whether the customer requires first-article inspection based on their standard terms).

Step 5: Quantify the Rework Loops

Every process has loops where work goes backward. A quote gets sent and the customer asks a question that requires re-pricing. A job hits the floor and the setup machinist discovers a discrepancy between the traveler and the drawing. An inspection report fails review and needs correction before the shipment can go out.

Track the frequency and cost of each rework loop. How often does each one occur? How much time does it consume when it does? What is the root cause? In many cases, rework traces back to information errors that happened three or four steps earlier in the process, not to the step where the rework occurs.

Step 6: Build the Current-State Map

Assemble your findings into a single visual document. A simple flowchart works. Boxes for steps, diamonds for decisions, arrows for flow, and colored highlights for wait states and rework loops. Include the time and system data for each step.

This current-state map should look messy. If it looks clean and linear, you have not captured enough detail. A real manufacturing workflow for quoting alone typically involves 15 to 25 discrete steps, 4 to 8 different systems, and 3 to 5 people. The map should reflect that complexity.

Step 7: Identify the Improvement Targets

With the current-state map complete, the improvement targets become visible. They fall into three categories:

1. Elimination. Steps that exist only because two systems do not talk to each other. Data that gets entered twice because no integration exists. Approvals that add time but not value.
2. Acceleration. Steps that take 30 minutes and could take 2 minutes with better data access. Searches through ERP records that could be pre-populated. Material cost lookups that could be automated.
3. Error reduction. Handoffs where information changes format and introduces transcription risk. Manual data entry points where a copy-paste or integration could eliminate the error opportunity.

Prioritize by impact and feasibility. The highest-value targets are usually the wait states and information retrieval steps, because those compress cycle time without changing how your people do their actual work. A value stream map builds on this same foundation with deeper time analysis.

Map first. Measure second. Fix third. Every process improvement that lasts starts with seeing the work clearly before changing it.