# Value Stream Mapping Value stream mapping maps how material and information move through a process, and separates value-adding time from the seven kinds of waste. **The question.** Where does time and material sit between the customer order and the finished item, and which of that is waste? Also known as: VSM. Also called material-and-information-flow mapping. The seven wastes are Ohno's muda. ## When to use it - Lead time is long, nobody can say where the days go, and each department can show that its own step is fast. - Improvement projects keep optimizing local steps, and work-in-process between them is growing. - A lean program is about to start, and the team needs one picture of door-to-door flow before picking tools. - Information flow, not only material, is suspected: forecasts, orders, and schedules that launch work too early. ## When to avoid it - The problem sits inside one step, such as a machine cycle or a single defect type. Map that step. A value-stream map will be too coarse. - There is no agreed product family. Mapping "everything we make" produces a drawing nobody can act on. - The client wants a wall mural as the deliverable. The map is a means to a future-state plan, not the product. - Demand is one-off project work with no repeating family. The method assumes a family that runs again and again. ## Overview Value stream mapping is a pencil-and-paper picture of how material and information move through one product family, from the customer back to the supplier. It separates the minutes of value-adding work from the days of inventory, waiting, and rework, using Ohno's seven wastes as the vocabulary. It supports the choice of which loop of the flow to change first, instead of speeding up every cell. The current-state map is only useful if a future state and a dated plan follow it. Most of the lead time will sit in queues, not in cycle time, which is why a flowchart of process steps misses the point. Map one family, walk the actual flow, and treat overproduction as the waste that usually creates the others. ## The seven wastes ### Transportation Moving material or information between processes that are separated in space. The move adds no value. It is often a symptom of layout and of batching. ### Inventory Material, or a queue of work, sitting between steps. Inventory hides problems, extends lead time, and is the easiest waste to count on a map. It is usually the largest block of time. ### Motion People walking, reaching, or searching to do the work. Motion is inside a process. Transportation is between processes. A spaghetti diagram shows motion. The value-stream map only hints at it. ### Waiting People, machines, or jobs idle because the previous step, a decision, or a part has not arrived. Waiting is often the other face of inventory. ### Overproduction Making more than the next process needs, or making it earlier. Ohno treated this as the worst waste, because it forces the others. Forecast-driven push and large batches are the usual sources. ### Overprocessing Extra operations, tighter tolerances than the customer needs, or inspections added because the process cannot be trusted. The customer does not pay for the extra work. ### Defects Scrap, rework, and the inspection built to catch them. On a map they show up as loops, extra boxes, and a complete-and-accurate rate well below 100 percent. ## What it is not - A process flowchart. A flowchart shows steps and decisions. A value-stream map shows material and information together, with times, inventories, and a lead-time bar. - A layout drawing or a spaghetti diagram. Those show movement in space. The value-stream map shows flow in time, from customer back to supplier. - A current-state drawing with no future state. Mapping that stops at "here is our waste" is a poster. ## The steps 1. **Pick one product family.** Group items that use roughly the same processes in the same order. Map that family, not the whole plant. Output is a family statement and the customer demand rate, usually as takt time. 2. **Draw the current state from the customer backward.** Walk the flow. Draw process boxes where material actually stops, data boxes for cycle time, changeover, uptime, and complete-and-accurate rate, inventory in days, and the information that tells each process what to make. Do it in pencil, on one page, with the people who run it. 3. **Mark the wastes and the lead-time bar.** Add value-adding time and calendar lead time on a bar under the map. Most of the bar will be inventory and waiting. Name which of the seven wastes dominate. Output is a current-state page the room recognizes. 4. **Draw a future state that removes the largest wastes.** Design flow around takt, a pacemaker, pull where continuous flow will not yet hold, and a leveling rule. The future state is a hypothesis, not a layout. Keep it on one page. 5. **Write a value-stream plan and review it.** Break the future state into loops with owners, dates, and a review cadence. Remap when a loop is done. The plan is the deliverable. The maps are how you argue it. ## In practice A components plant quotes 21 days from order to ship on its highest-volume family. Each cell reports a cycle time of a few minutes. The consultant walks the family from the shipping dock backward with the supervisors, drawing on paper. Process boxes show 48 minutes of value-adding work. Between the boxes sit 19 days of inventory: a six-day raw-material supermarket, a five-day queue in front of plating (a batch process run twice a week), and eight days of finished goods built to a monthly forecast. Information flow is a weekly MRP drop that launches two weeks of work at every process. Overproduction and waiting, not slow operators, account for the 21 days. The future-state sketch cuts the MRP launches to a pacemaker at the last cell, sized to daily demand, and changes plating from two large batches a week to a daily slot for this family. The implementation plan has three loops over twelve weeks: daily plating slots first, then a supermarket sized in days of demand, then the pacemaker. A SIPOC would have named the steps. It would not have shown that 48 minutes of work were sitting inside 21 days. ## What you leave with - A named product family and a demand rate. - A current-state map with process data, inventories in days, and information flow. - A lead-time bar that splits value-adding time from the rest. - A future-state sketch and a looped implementation plan with owners and dates. - A short list of the wastes that the plan actually attacks. ## Failure points ### Mapping everything One drawing tries to hold every product and every exception. The page is unreadable, and no future state can be drawn. Pick a family. ### Drawn in a conference room The map records the documented process. Walking the floor finds the queues, the unofficial supermarket, and the rework loop that the procedure omitted. ### Current state as the deliverable The team laminates the as-is map and stops. Without a future state and a plan, nothing in the flow changes. ### Future state as a wish list The to-be map assumes new equipment, a new layout, and a new ERP in one step. A future state should be reachable in months with a few loops, not a capital program. ### Local kaizen on a non-constraint A cell is tidied and sped up, and the inventory in front of the real bottleneck grows. The lead-time bar does not move. Tie the loops to the constraint in the stream. The failures above are usually [Framing](https://usefulconsultingframeworks.com/your-edge/#framing): the wrong problem was chosen before this framework was applied. ## Origin Taiichi Ohno classified non-value-adding work in the Toyota production system as muda, the seven wastes. Overproduction was the one he treated as the source of the others, because extra output creates inventory, motion, and hiding places for defects. The mapping method that makes those wastes visible dock-to-dock was written for Western plants by Mike Rother and John Shook in Learning to See, published by the Lean Enterprise Institute in 1998. The workbook teaches a current-state map, a future-state map, and an implementation plan, using pencil and paper and a standard set of icons. The method is not licensed. The icons and the Learning to See workbook are copyrighted by the Lean Enterprise Institute. This note describes the practice in its own words and does not reproduce the icon set. Later writers added an eighth waste, unused talent. That is a later overlay, not Ohno's list. ## Further reading - [Value Stream Mapping, Lean Enterprise Institute lexicon](https://www.lean.org/lexicon-terms/value-stream-mapping/). LEI's free definition of the map, the data boxes, and the point of current-state mapping. Matches the vocabulary in Learning to See. - [Learning to See](https://www.lean.org/store/book/learning-to-see/). Mike Rother and John Shook, Lean Enterprise Institute, 1998. The workbook that taught the method. A paid book, with sample chapters on the same site. ## Related frameworks - [SIPOC](https://usefulconsultingframeworks.com/frameworks/sipoc/): A one-page boundary before you walk the floor. Use it first when people do not agree where the process starts and stops. - [PDCA](https://usefulconsultingframeworks.com/frameworks/pdca/): The cycle to run on each implementation loop once the future state is drawn. - [Theory of Constraints](https://usefulconsultingframeworks.com/frameworks/theory-of-constraints/): Finds the one constraint the future state should protect. Pick it when speeding up every step would only move the pile. - [5S](https://usefulconsultingframeworks.com/frameworks/5s/): Makes a workplace orderly enough to see queues, missing parts, and extra motion. It often comes before a reliable walk of the stream. - [Lean Startup](https://usefulconsultingframeworks.com/frameworks/lean-startup/): A discovery cycle for whether customers want something. Pick it when the uncertainty is demand, not waste in a process that already ships. Last updated October 20, 2026.