VSM Works

What is value stream mapping?

Value stream mapping is a lean method for drawing every step a product takes from raw material to customer — the material flow, the information flow that triggers it, and the time each step consumes — on a single page. The purpose is not the drawing. It is the timeline underneath it: the comparison between how long the work actually takes and how long the product spends waiting, which in most plants turns out to be a ratio of minutes to weeks.

What it is for

The method comes from Toyota, where it was called material and information flow mapping, and reached a wider audience through Rother & Shook's Learning to See in 1999. The name it was given there says what it is: the point is to see a flow that nobody in the plant can see, because each person can only see their own station.

A plant that measures itself station by station will find every station busy and conclude it is running well. The value stream map is the instrument that contradicts this. It puts the processing time and the waiting time on the same axis, and the resulting comparison — processing time against production lead time — is routinely something like 188 seconds of work inside 23.6 days of lead time. That figure is from the worked example in Learning to See, and it is not an unusual one. Under 1% of the time a product spends in a plant is typically spent being worked on.

Everything else the method does follows from making that visible. You cannot argue for removing a buffer that nobody has drawn.

How it differs from a process map or flowchart

This is the distinction most often collapsed, including by tools that offer "value stream map" as a shape library. Three things separate them:

What a map is made of

A value stream map is built from a small, fixed vocabulary — which is a large part of why it travels: two people from different companies can read each other's maps.

Current state, then future state

A value stream map is really two maps and a plan. The current state is what the plant does today, drawn from observation rather than from the routing file. The future state is a design for what it should do, drawn against the same customer demand. The plan is how one becomes the other.

The order is not negotiable, and the commonest way the exercise fails is skipping the first map because everybody already knows what the problems are. They do not: the current state routinely contradicts what the plant believes about itself, and a future state designed against a belief is a redesign of an imaginary factory.

Nor is a future state a wish list. It is constrained by takt time — the rhythm the customer's demand sets — and every design decision in it is answerable to whether the resulting flow can hold that rhythm.

The numbers a map produces

The drawing is the input. These are the outputs, and each of them is defined with its formula in the glossary:

When not to map

The method has a domain, and applying it outside that domain produces a diagram that costs three days and answers nothing. Do not start a value stream map when:

Common questions

Is value stream mapping the same as process mapping?

No. A process map shows the sequence of steps. A value stream map adds the information flow that triggers each step and a timeline that separates work time from waiting time — and it is those two additions that turn a picture into an analysis. A value stream map also covers one product family door to door, where a process map can cover any scope.

How long does it take to draw one?

A current-state map for a single product family is typically a matter of days rather than weeks: one to two days walking the floor and collecting data, and a day to draw and check it. The constraint is almost never drawing speed — it is getting real cycle times and real inventory counts, which requires being on the floor with a stopwatch rather than in the ERP system.

Do I need software to do it?

No. The first map should arguably be pencil and paper, and Learning to See recommends exactly that — the walking and counting are the exercise. Software earns its place afterwards, when the map has to be recalculated, compared against a future state, and turned into a report that agrees with itself.

What data do I need to collect?

Per process: cycle time, changeover time, uptime, number of operators, shift pattern, batch size and scrap or rework rate. Between processes: how much inventory is sitting there, counted rather than looked up. Plus customer demand and the working calendar, which are what takt time is computed from.

Does value stream mapping only apply to manufacturing?

It began there and its vocabulary shows it, but the method extends to any repeating process with handoffs and queues — order processing, claims handling, clinical pathways, software delivery. Martin & Osterling's Value Stream Mapping is the standard reference for that extension. The mechanics are the same; the inventory triangles become queues of work rather than piles of parts.

Can I skip the current state and design the future state directly?

This is the most common way the exercise fails. The current state routinely contradicts what a plant believes about itself — inventory is somewhere nobody expected, or the constraint is not the station everyone blames. A future state designed against a belief instead of an observation is a redesign of a factory that does not exist.

Sources


Everything above can be done on paper, and for a first map it probably should be — the walking and the counting are the exercise, not the drawing. What paper does not do is the arithmetic, and the arithmetic is where a map quietly stops agreeing with the report made from it.

VSM Works is a value stream mapping tool built around that problem: one calculation engine produces the on-screen metrics, the A3 print, the PDF, the workbook and the image, so a figure in a presentation cannot disagree with the map it came from. It runs in the browser, needs no account to try, and works in English and Turkish. Its methodology is published in full, and its arithmetic is checked against the totals Rother & Shook published.

Open the editor