A quote that takes nine working days usually contains about an hour of real work. The rest is waiting: in a shared inbox, in someone’s to-do list, in an approval folder. Nobody sees that waiting, because in an office it does not pile up on the floor. It hides in email.
Value stream mapping makes the waiting visible and puts a number on it. You follow one request from start to finish, write down every step, measure how long the work takes and how long the work sits still, and the gap between those two numbers tells you where to start. The method comes from Toyota and was written down for the rest of the world by Mike Rother and John Shook in the workbook Learning to See, published by the Lean Enterprise Institute. This guide uses the definitions of the Lean Enterprise Institute and ASQ, adds a worked office example with real calculations, and shows how a value stream map becomes an automation roadmap. That last part is how Viralistic works: we map a process before we build anything in n8n, because automating a step that should not exist only produces waste faster. If Lean is new to you, start with our guide to Lean management; if you already know which process hurts, an automation specialist can map it with you.
What is value stream mapping?
Value stream mapping is a visual Lean method that diagrams every step in the material and information flow needed to bring a product or service from order to delivery, with process data under each step. Its purpose is to see where value is created and where time, effort and errors are lost, so you know what to improve first.
A value stream is all the actions, value-creating and not, that it takes to deliver to the customer. The Lean Enterprise Institute notes that Toyota developed the tool under a plainer name: the material and information flow diagram, a working part of the Toyota Production System. That name says what makes a value stream map different from a flowchart. It draws two flows on one page. The material flow is the thing moving towards the customer: a part in a factory, a request, a file or a quote in an office. The information flow is what tells each step what to do next: a production schedule, an email, a CRM notification.
Under every step sits a data box with real measurements, and along the bottom runs a timeline that adds up waiting time and working time. That timeline is the reason the method works. Most teams know their process is slow. Few know that their quote spends 97% of its life waiting.
A note on the terms: a value stream map is the drawing itself, value stream mapping is the activity of making it, and VSM is used for both.
Current state, future state and ideal state
A value stream map comes in three versions. The current state map shows how the process runs today, with every wait and every rework loop. The future state map shows how it should run within realistic limits, as the target for the coming months. Some teams add an ideal state map: how the process would look with no constraints at all, used to stretch ambition rather than as a plan.
The current state is always first, and it has to be honest. You draw what happens, not what the procedure manual says should happen. The future state is designed from it: you remove steps, shorten waits and mark each change with a kaizen burst. The difference between the two maps becomes your improvement plan, and ASQ calls that plan the most important part of the whole exercise. A beautiful current state map with no future state is wallpaper.
Value stream mapping symbols
Value stream maps use a small set of standard symbols, so anyone who has learned the method can read anyone else’s map. The table shows the ones you need most, with what each one means in an office process. The Lean Enterprise Institute offers the full icon set for download from its value stream mapping page.
| Symbol | What it shows | In an office value stream |
|---|---|---|
| Factory outline | The customer (top right) or the supplier (top left) | The client who asks; the partner or system that supplies input |
| Process box | One step where the work flows | Draft quote, approve, send invoice |
| Data box | The measurements under a process box | Process time, waiting time, %C&A, people involved |
| Triangle with an I | Inventory waiting between two steps | Unread emails, tickets in a queue, files in an approval folder |
| Striped arrow | Push: work sent on whether the next step is ready or not | “Can you pick this up?” emails |
| Thin straight arrow | Manual information flow | A phone call, a paper form, a word at the coffee machine |
| Lightning arrow | Electronic information flow | Email, a CRM notification, a webhook |
| Supermarket and kanban | A controlled buffer the next step pulls from | Pre-approved quote modules, standard templates |
| FIFO lane | First in, first out, with a maximum | A request queue handled in order, with a cap on open items |
| Operator icon | The people working at a step | Number of people who touch the request |
| Kaizen burst | An improvement to make | Marked on the future state map |
| Timeline | Waiting time on the upper line, process time on the lower line | The totals that show lead time against working time |
Do not let the symbols slow you down. Draw the first version with a pen and sticky notes, with the people who do the work, and tidy it up in a tool later.
Value stream mapping metrics: lead time, cycle time, process time and value-added ratio
A value stream map measures time and quality at every step, then adds them up. Lead time is the total time a single request needs to go through the whole stream. Process time is the part of that in which someone actually works on it. The value-added ratio divides the time the customer would pay for by the lead time, and in office processes it is often shockingly low.
| Metric | What it means | How to calculate it |
|---|---|---|
| Lead time (L/T) | Time for one item to move through a step or the whole stream, including all waiting | Timestamp at the end minus timestamp at the start |
| Process time (P/T) | Hands-on time spent working on one item at a step | Measured or observed per step |
| Cycle time (C/T) | How often a step completes an item; the time to run through all its work before repeating | Observed per step |
| Value-added time | The part of process time that changes the item in a way the customer would pay for | Judged per step, from the customer’s point of view |
| Value-added ratio | The share of lead time that adds value; also called process cycle efficiency | Value-added time ÷ total lead time × 100 |
| Activity ratio | The share of lead time in which anyone is working on the item at all | Total process time ÷ total lead time × 100 |
| %C&A | Percent complete and accurate: how often a step receives work it can use without correction | Share of items the next step accepts without rework |
| Rolled %C&A | How often an item passes through the whole stream without a single correction | All %C&A values multiplied |
| Takt time | The pace of customer demand | Available working time ÷ customer demand |
Lead time vs cycle time vs process time
People mix these three up, and the mix-up hides the problem. Process time is the work. Lead time is the work plus all the waiting around it. Cycle time, in the Lean Enterprise Institute’s definition, is how long a step takes to go through its work before it starts on the next item, which tells you about capacity rather than about the customer’s wait. A quote can have a process time of 60 minutes, a cycle time of an hour and a lead time of four days. The customer only feels the four days.
Takt time, %C&A and the rolled figure
Takt time comes from manufacturing but works anywhere demand is steady. The Lean Enterprise Institute’s own takt time example: a factory that runs 480 minutes a day for customers who want 240 widgets a day has a takt time of two minutes. An office that receives 16 quote requests in a 480-minute day has a takt time of 30 minutes.
%C&A is the quality metric that matters most in office work, because errors in information do not look like scrap. A missing phone number, a wrong price, an outdated template: each sends the work back. Karen Martin and Mike Osterling, whose book on value stream mapping focuses on office and service processes, multiply the %C&A of every step into a rolled figure. Five steps that each get it right 90% of the time only produce a clean result 59% of the time.
How to create a value stream map in 7 steps
Creating a value stream map takes seven steps: choose one value stream, assemble the people who do the work, walk the process, draw the boxes and flows, fill the data boxes with real measurements, draw the timeline, and design the future state with a plan. NIST describes the same core loop in four steps: pick the value stream, map the current state, design the future state, implement.
- Choose one value stream. Pick one product or service family with a clear start and end, for example “from enquiry to quote in the client’s inbox”. The Lean Enterprise Institute advises one person to lead the mapping effort personally. Choose the stream with the most waiting, the most complaints or the most volume.
- Bring the people who do the work. ASQ suggests a cross-functional team of about ten for a facility-level map. An office stream usually needs one person per handover, plus whoever owns the result. Managers alone will draw the process as it was designed, not as it runs.
- Walk the process. Lean calls this going to the gemba, the place where the work happens. In digital work, follow one real request through the inbox, the CRM, the documents and the approvals, starting at the customer and working back.
- Draw the process boxes and both flows. One box per step where the work stops and changes hands. Add the information flow above: who tells whom to do what, by email, chat or phone.
- Fill the data boxes with measurements, not opinions. Process time, waiting time before the step, %C&A and the number of people. In an office, timestamps from email, the CRM or the ticket system beat stopwatches and guesses. Record the spread too: a step that usually takes ten minutes and sometimes two days hides inside an average.
- Draw the timeline and do the sums. Add up waiting time and process time, then calculate lead time, value-added ratio and rolled %C&A.
- Design the future state and write the plan. Mark kaizen bursts where the waits and the rework are biggest, pick two or three, and give each one an owner and a date. Then check the result, which is where the PDCA cycle takes over.
The map is finished when the team agrees it shows reality. It does not need to be pretty. The Lean Enterprise Institute puts it plainly: a perfectly drawn current state map is not the point; understanding the flow is.
Value stream mapping example: a quote process in an office
Here is a worked value stream mapping example from office work: the quote process of a premium service firm, from the moment an enquiry arrives to the moment the quote lands in the client’s inbox. The firm and its numbers are illustrative, but the pattern is the one most offices recognise. A working day is 480 minutes.
The current state
| Step | Process time | Waiting before the step | %C&A |
|---|---|---|---|
| 1. Enquiry read in the shared inbox and forwarded | 5 min | 4 hours | 95% |
| 2. Enquiry re-typed into the CRM | 10 min | 1 day | 90% |
| 3. Intake call scheduled by email and held | 45 min | 2 days | 100% |
| 4. Quote drafted in Word from the CRM and price list | 60 min | 3 days | 80% |
| 5. Owner approves the quote | 10 min | 2 days | 85% |
| 6. Quote exported to PDF, sent and logged | 10 min | 4 hours | 95% |
| Total | 140 min | 4,320 min (9 days) |
What the current state map shows
The lead time is 4,320 minutes of waiting plus 140 minutes of work: 4,460 minutes, or 9.3 working days. Of those 140 minutes of work, only two parts change anything the client would pay for: the 30-minute intake conversation and the 30 minutes spent tailoring the scope. That gives:
- Value-added ratio: 60 ÷ 4,460 × 100 = 1.3%
- Activity ratio: 140 ÷ 4,460 × 100 = 3.1%
- Rolled %C&A: 0.95 × 0.90 × 1.00 × 0.80 × 0.85 × 0.95 = 55%
So for 97% of its life, nobody touches the quote, and almost half of all quotes are sent back at least once because of a typo, a wrong price or a missing detail. Working harder at any single step changes almost nothing. The waits sit between the steps, which is exactly where nobody is responsible.
The future state
The team marks four kaizen bursts: the CRM re-typing, the scheduling ping-pong, the drafting backlog and the approval queue. Two of the four changes need no software at all.
| Step | Process time | Waiting before the step | %C&A |
|---|---|---|---|
| 1. Website form writes the enquiry to the CRM and sends a booking link (n8n) | 0 min | none | 100% |
| 2. Intake call in a slot the client picked | 30 min | 1 day | 100% |
| 3. Draft prepared from standard quote modules and the call notes; consultant tailors scope | 30 min | 4 hours | 95% |
| 4. Owner approval only above a set amount; standard quotes skip it | 0 min | none | |
| 5. Quote sent and logged in one click | 2 min | none | 100% |
| Total | 62 min | 720 min (1.5 days) |
The lead time drops from 9.3 to 1.6 working days (782 minutes). The value-added ratio rises from 1.3% to 7.7%, and the rolled %C&A from 55% to 95%. Look at what did not change: the 60 minutes of value-added work. The client still gets a real conversation and a scope written by a person. Everything that disappeared was waiting, re-typing and correcting.
The approval rule alone removes two days of waiting, and it costs one sentence in a policy, not a workflow. Standard quote modules, a Lean supermarket in office form, remove most of the wrong prices before any software is involved. Only then does automation earn its place, on the steps that are left: the form that fills the CRM, the draft prepared from structured call notes, the one-click send. That order is the whole point.
Value stream mapping vs process mapping vs swimlanes
A value stream map shows where the time goes across a whole stream, with data. A process map, or flowchart, shows the steps and decisions of one process in detail. A swimlane diagram shows the same steps split by role, so the handovers become visible. They answer different questions, and in practice you often need two of them.
| Value stream map | Process map (flowchart) | Swimlane diagram | |
|---|---|---|---|
| Main question | Where does the time go? | What are the steps and decisions? | Who does what, and where are the handovers? |
| Scope | The whole stream, request to delivery | One process, often in fine detail | One process, split by role, team or system |
| Data | Always: times, waits, %C&A | Rarely | Sometimes |
| Information flow | Drawn separately | Mixed into the steps | Visible as lines crossing lanes |
| Best for | Cutting lead time, choosing what to fix first | Work instructions, decision logic | Fixing handovers and ownership |
| Role in automation | The roadmap: which steps to remove or automate | The logic of one workflow | Owners, approvals and human checkpoints |
The Lean Enterprise Institute sums up the difference: taking a value stream perspective means improving the whole, not optimising the parts. Two other tools sit nearby. A SIPOC (suppliers, inputs, process, outputs, customers) gives a one-page overview before you map. Process mining reconstructs a process from event logs in your systems; it is a good source of timestamps for the data boxes, but it only sees what the systems record, not the email or the phone call in between.
Is value stream mapping a Six Sigma tool? It started as a Lean tool, and Lean Six Sigma programmes adopted it, often in the Define or Measure phase of DMAIC, where it shows the team which problem deserves a closer statistical look.
From value stream map to automation roadmap
A future state map is the best automation brief you can write. It shows which steps should disappear, which should be standardised, and which steps remain that software can run better than people. Automating from the current state instead locks the waits, the re-typing and the rework into code.
That is the stance we take at Viralistic, and it is not a slogan: never automate a broken process. Automating waste produces waste faster, with fewer people watching. Every step on the future state map gets one of five labels, in this order:
- Remove. The step adds nothing the customer would pay for and no rule requires it. Most approval loops and status updates land here. These are the eight wastes of Lean in office form: waiting, re-entering data, searching for files, reports nobody reads.
- Simplify. The step is needed but messy. Standardise it first: one template, one price list, required fields, a clear definition of done.
- Automate. The step follows a rule you can write down: the same input always gives the same output. Build it as a deterministic workflow in n8n.
- Assist. The step needs judgement: reading free text, summarising a call, drafting a scope. Let an AI model such as Claude prepare it, with a person checking before anything leaves the company.
- Keep human. The step is the value: the intake conversation, the advice, the relationship.
Then rank the automate and assist steps by waiting time multiplied by volume, and start with one to five workflows, not twenty. The future state map has already given you the metrics: the lead time and rolled %C&A you expect after the change. Those become the targets you measure the workflows against. Toyota had a word for the idea behind good automation, jidoka, automation with a human touch: the machine stops and flags a problem, and people improve the process. A well-built workflow does the same with error handling and alerts, and the team keeps improving the map in small kaizen steps.
How Viralistic maps before it builds in n8n
At Viralistic, every automation project starts with a map, not with a tool. We sit down with the person who does the work and write down every step, what starts it, which systems it touches, what is decided and on what basis, and what goes wrong regularly. Then we mark each step as rule-based or judgement, keep as much as possible deterministic, and decide in advance where a person checks the work. The method is written out step by step in our guide to turning a business process into an agentic workflow.
From there we agree on the value in a success plan, build one to five workflows with n8n, Claude and Notion, and roll them out in stages: shadow mode first, then assisted, then automated for the cases that proved reliable. Error handling, logging and access rights are built in from day one. Read more about our value-first automation approach and what to expect from workflow automation services.
The method was not invented for this post. Marrallisa Kreijkes, founder of Viralistic in Amsterdam and an official n8n Ambassador, taught it in her DataCamp webinar How to turn a business process into an agentic workflow. She trained Volkswagen Pon Financial Services on getting found in AI, spoke on the main stage of n8n’s conference In The Loop in Berlin in October 2026, gave a lecture at Anthropic, the company behind Claude, and ran an n8n workshop at Maastricht University with the student associations MAIN and TCM. Teams at DataCamp, Volkswagen, n8n and Anthropic learned from her. When you work with Viralistic, you work with the person they brought in to teach. More about her talks and trainings.
Value stream mapping is the cheapest step in any automation project and the one that saves the most money, because it stops you from building the wrong thing. If you have a process that takes days for an hour of work, work with an automation specialist who maps it first, or read the Lean management guide to see how VSM fits the bigger picture.
Frequently asked questions
What are the four steps of value stream mapping?
The four classic steps are: identify the value stream to map, draw the current state map with real data, design the future state map, and implement the improvements with a plan. NIST and ASQ describe the same sequence. In practice, step two takes the most time, and step four is the one teams skip, which is why many maps never change anything.
What are the three types of value stream maps?
The three types are the current state map, the future state map and the ideal state map. The current state shows how the process runs today, with every wait and rework loop. The future state shows the realistic target the team commits to next. The ideal state shows the process without any constraints and is used to stretch thinking, not as a plan.
Is value stream mapping a Six Sigma tool?
Value stream mapping is a Lean tool, developed at Toyota as the material and information flow diagram. Lean Six Sigma programmes adopted it, often in the Define or Measure phase of DMAIC, to show where time and errors pile up before the team chooses which problem to analyse statistically. You do not need any belt to draw a useful one.
What is the best tool for value stream mapping?
Start with paper, sticky notes and a pen, together with the people who do the work, because the conversation matters more than the drawing. Afterwards, digital whiteboards such as Miro or Lucidchart and Microsoft Visio include standard VSM symbols. The best tool is the one your whole team can see and edit, not the one with the most icons.
Can you make a value stream map in Excel?
Yes. Put the process steps in columns, with rows for process time, waiting time, %C&A and people, and let formulas calculate lead time, value-added ratio and rolled %C&A. Excel is weak at drawing the information flow, so many teams sketch the map on a whiteboard and keep the numbers in a spreadsheet next to it.
What is an example of a value stream?
A value stream is every action needed to deliver one thing to a customer. Examples are an order from purchase to delivery, a quote from enquiry to the client’s inbox, an invoice from finished work to payment, and a new employee from signed contract to first productive day. Each one crosses several people and systems, which is where the waiting hides.
How long does value stream mapping take?
A first current state map of one office process takes a day or two, including walking the process and collecting timestamps. ASQ describes a kaizen event of two to five days to produce the current state, the future state and a draft plan. Implementing the future state then takes weeks to months, depending on how many changes need software.
What are the limitations of value stream mapping?
A value stream map is a snapshot, so it ages as the process changes. It works poorly for processes that are rarely repeated or still unstable, and it hides variation if you only record averages. It also says where the problems are, not why. For root causes you still need tools such as the 5 whys or a fishbone diagram.
Should you automate the current state or the future state?
Always the future state. Automating the current state builds today’s waits, double entry and rework into software, so they run faster and are harder to change. Remove and simplify first, then automate the rule-based steps that are left and add AI with a human check where judgement is needed. That order keeps the automation small and reliable.
Written by
Founder of Viralistic and official n8n Ambassador
Marrallisa leads strategy, design, search and automation for premium brands from Amsterdam. She has trained teams at Volkswagen and speaks at n8n and Anthropic events. Want to talk through your situation? Book a call with her directly.