Implementation Guide
Value Stream Mapping.
Value stream mapping (VSM) is a Lean method for drawing how material and information actually flow from order to delivery for one product family, so you can see where time and value are lost. A good mapping week produces three things: a current-state map the team believes, a future-state design, and a prioritized improvement plan. The map is not the deliverable. The improvement plan is.
What VSM is, and what it is for
A value stream is every step, value-adding or not, that one product family passes through between order and delivery, plus the information flow that triggers each step. Value stream mapping draws that whole system on one page: processes with measured data, inventory between them counted in days, the scheduling logic on top, and a timeline at the bottom separating time spent working from time spent waiting.
The method comes from Toyota, where it was called material and information flow mapping, and was popularized by Mike Rother and John Shook in Learning to See. The name change matters less than the two words most maps forget: information flow. Most delivery problems are not caused by slow machines. They are caused by how work is scheduled, batched and pushed between machines.
The typical first discovery is uncomfortable and useful: a part that takes 42 minutes of actual processing spends two and a half weeks in the building. Almost all lead time is waiting, and waiting is caused by decisions that a map makes visible: batch sizes, changeovers, push scheduling, and inventory nobody designed.
Where VSM sits in the transformation roadmap
VSM is an early diagnostic and design tool, not an isolated Lean technique. On the TeamGuru deployment roadmap it belongs to the Diagnose stage: it runs after a first current-state assessment has framed the business problem, in parallel with baseline KPIs, and it feeds directly into future-state design and the priorities deployed through strategy deployment.
- Before Current-State Assessment
- You are here Value Stream Mapping
- In parallel Baseline KPIs
- After Future State Design
- After Strategy Deployment
When to use it, and when not to
Map when you need to decide where improvement effort should go: lead times are too long, delivery is unreliable, inventory keeps growing, or improvement projects keep landing wherever the loudest voice points. The map replaces opinion with a shared, measured picture, and the future state turns that picture into a small number of high-leverage projects.
Do not start with VSM in these situations:
- An acute safety, quality or customer crisis is running. Contain it first; a mapping week does not stop a line-down situation.
- Nobody in the room has the authority to change the flow. A map without a sponsor becomes wall art.
- The goal is documentation, certification or a slide for a steering committee. VSM is a decision tool; if no decision is pending, save the week.
- You cannot name a product family. Map one family that matters, not the whole plant at once.
Key terms and metrics
These are the numbers a mapping team measures and argues about. If a term on the map cannot be traced to something someone counted or timed at the process, it does not belong on the map.
| Term | What it means on the map |
|---|---|
| Value stream | Every step, value-adding or not, required to take one product family from order to delivery, including the information that triggers each step. |
| Current-state map | A drawing of how material and information actually flow today, with measured data, not how the ERP or the process documentation says they flow. |
| Future-state map | A design of how the same stream should flow after improvement, typically 6 to 12 months out. |
| Takt time | Available working time divided by customer demand. The rhythm the stream must produce at to meet demand exactly. Example: 54,000 seconds of daily working time for 456 units gives a takt of 118 seconds. |
| Cycle time (C/T) | How often one finished unit actually leaves a process, measured at the process. |
| Changeover time (C/O) | Time from the last good unit of one product to the first good unit of the next. Long changeovers force large batches, and large batches create inventory. |
| Lead time | Total time one unit spends in the stream, queues included. Read off the map's timeline ladder in days. |
| Process cycle efficiency | Processing time divided by lead time. In most unmapped streams this is below 1 percent, which is normal, not shameful. |
| Supermarket and kanban | A controlled inventory buffer with a signal that triggers replenishment only when something is consumed. The standard way to replace push scheduling between processes that cannot yet flow. |
| Pacemaker | The single process that receives the production schedule. Everything upstream is pulled; everything downstream flows. One schedule point instead of five competing ones. |
| Kaizen burst | A starburst symbol marking where the future state requires an improvement project, for example a changeover reduction. Bursts become the improvement plan. |
The standard symbols, in one table
| Process box | One processing step with its data box (C/T, C/O, uptime, operators, shifts) |
|---|---|
| Data box | The measured facts about a process, collected at the process, not from reports |
| Inventory triangle | A count of material waiting between processes, converted to days of demand |
| Push arrow (striped) | Material moved on a schedule regardless of whether the next process needs it |
| Supermarket | A controlled buffer that the downstream process withdraws from |
| Kanban signal | The instruction to produce or move, triggered by consumption |
| Truck / shipment | Deliveries in and out, with frequency |
| Information flow | Straight line for paper or verbal, zigzag for electronic (MRP, EDI, email) |
| Timeline ladder | The bottom line of the map: waiting days on top, processing time below. This is where the lead time comes from |
| Kaizen burst | A required improvement, and one future row of your project list |
How to run a mapping week
A first map does not take months. It takes one prepared team and about four days, and it ends with a plan presented to site leadership, not with a drawing filed in a folder. This agenda has survived contact with real plants; adapt the days, keep the sequence.
| When | What happens |
|---|---|
| Before the week | Pick one product family that carries volume or margin. Name the team: value stream manager, supervisors and operators from the stream, one planner, one person with authority to change the flow. Block the room and the calendars. |
| Day 1 morning | Walk the stream door to door, starting at shipping and moving upstream. No mapping yet, just seeing and asking. |
| Day 1 afternoon | Walk again with stopwatches and count sheets. Collect cycle times, changeover times, uptime, operator counts, and count every pile of inventory. |
| Day 2 | Draw the current state on paper or a whiteboard, together. Material flow left to right, information flow on top, timeline ladder at the bottom. Argue about the numbers until the people who do the work accept them. |
| Day 3 morning | Analyze: compare cycle times to takt, mark every push arrow, circle the inventory pools, and ask what makes each one necessary. |
| Day 3 afternoon | Design the future state: where can flow replace queues, where must supermarkets cap inventory, which single process gets the schedule, and which changeovers must shrink to allow smaller batches. Mark every required improvement as a kaizen burst. |
| Day 4 | Turn bursts into an improvement plan: group them into loops, name an owner and a measurable target for each, and sequence the first 90 days. Present to the site leadership team the same day. |
Rules that keep the week honest
- Walk the flow before drawing it, and walk upstream, from shipping toward raw material, so you see the stream the way a customer order experiences it in reverse.
- Measure at the process. Cycle times from the ERP, capacity studies or memory are the fastest way to draw a fiction.
- Count all inventory, including the pallets nobody wants to talk about, and convert counts to days of demand so they land on the timeline.
- Draw by hand first. The point of the current state is a shared argument that ends in agreement, and a projector kills the argument.
- Design the future state in the same week, while the discomfort is fresh. A current-state map without a future state is a complaint, not a plan.
The five future-state questions
Design the future state by answering these in order. Every yes and no becomes a line on the map, and every gap becomes a kaizen burst.
- What is the takt time, and which processes cannot meet it today?
- Where can processes flow one piece at a time with no inventory between them?
- Where flow is not yet possible, where do supermarkets and kanban cap the inventory?
- Which single process becomes the pacemaker that receives the schedule?
- Which improvements, changeovers, uptime, layout, must happen to make the design real?
Worked example, with the reasoning
The numbers below are illustrative, but internally consistent, for the same 450-person components manufacturer used across the transformation roadmap: demand of 456 units per day across cutting, machining, welding and assembly, with a weekly MRP schedule pushed to every process. The mapping team measured this:
| Metric | Current state | Future state (12 months) |
|---|---|---|
| Lead time | 18.4 days | 7.2 days |
| Processing time | 42 min | 38 min |
| Work in process | 8,400 pcs | 3,300 pcs |
| Changeover (machining) | 47 min | 18 min |
| Schedule points | 4 processes | 1 pacemaker |
| Process cycle efficiency | 0.25% | 0.59% |
Current state
Four schedule points, push between every process, inventory pooling wherever batches collide.
Future state, 12 months
One schedule point at the pacemaker, pull everywhere else, inventory capped in supermarkets.
Read the first two rows together, because they carry the whole lesson: lead time falls by 61 percent while processing time barely moves. Nobody works faster in the future state. The stream simply stops waiting. With 456 units a day of demand, 8,400 pieces of WIP is 18.4 days of lead time by definition; cutting WIP to 3,300 pieces is what cuts lead time to 7.2 days. The design decisions that get there:
Schedule only final assembly and pull everything upstream
Why: Four separately scheduled processes meant four competing priorities and expediting between them. One pacemaker removes the argument about what to run next everywhere else.
Run a changeover reduction project on machining (47 to 18 minutes)
Why: The 47-minute changeover forced batches of roughly three days of demand. Below 20 minutes, batches can shrink by two thirds, and the 6-day inventory pool in front of machining shrinks with them.
Put supermarkets with kanban between cutting, machining and assembly
Why: The stream cannot flow piece by piece yet, so the buffers stay, but capped and self-managing. Inventory becomes a designed quantity instead of an accident.
Merge welding into the assembly cell
Why: Welding ran at half of takt and fed a 2.9-day queue. Moving it into the cell removes a handoff, a queue, and a schedule point at the cost of one layout change.
The output of the week was not the map. It was four projects with owners, targets and a 90-day sequence, plus two KPIs, lead time and changeover minutes, added to the plant's baseline so progress is visible weekly.
After the map: making it stick
Most value stream maps die in the two weeks after the workshop, and they die the same way: the bursts never become owned projects, the projects never get a review rhythm, and six months later someone finds the map behind a cabinet. What follows the map is a management chain, and every link needs an owner:
Current state → future state → gaps → projects → actions → KPIs → management review.
The first half of that chain is VSM. The second half is where a management system takes over, and where TeamGuru earns its place: the future-state map and its assumptions live in Documents next to the work instead of on a wall, each kaizen burst becomes a project with owned, dated actions, and lead time and changeover join the KPI baseline so the monthly review tracks the stream, not the memory of a workshop.
On the roadmap, the next moves after VSM are future-state design for the priority streams and strategy deployment, which turns the biggest gaps into the plant's few measurable priorities.
Common mistakes
What bad looks like
- The map is drawn from ERP data in an office, then presented to the floor
- Every process in the plant is mapped at once, and nothing changes
- The current state is finished; the future state is postponed to a later workshop that never happens
- The map is beautiful, laminated, and six months out of date
- Bursts have no owners, no dates and no KPI that would show progress
What good looks like
- One product family, one team, four days, one improvement plan
- Numbers measured at the process and accepted by the people who run it
- A future state dated 6 to 12 months out, with a takt everyone can quote
- Every kaizen burst is a project with an owner and a measurable target
- Lead time and changeover on the KPI baseline, reviewed monthly