Your Training Partner
Techniques Toolbox
Current-state VSM of a health insurer reimbursing an outpatient treatment invoice: five processing steps preceded by queues, a sawtooth timeline setting 20 working hours of waiting against 23 min of processing, kaizen bursts on three queues, cycle efficiency 1.9%.

Value Stream Mapping (VSM)

On a single page, value stream mapping draws a snapshot of every step that carries a product or a service to the customer. The map shows two flows together: the material and work flow that turns the request into a delivery, then the information flow that triggers each step. At each step it records the processing time and the wait time spent in a queue. Beneath the row of boxes, a timeline sums the processing times and the wait times separately. Their ratio shows the share of the total lead time in which the work moves forward.

Goal

Value stream mapping gives a cross-functional team a shared and evidence-based picture of an end-to-end flow. The team sees where time and value are lost, agrees on the waste and designs a leaner future state. It then plans the gains and measures them. The map addresses a question that local measures leave unanswered: why does a simple request take days to complete when the work itself takes a few minutes? It compares the time in which the request moves forward with the time it waits between steps.

The deliverable comprises a current-state map, a future-state map and the list of improvements that lead from one to the other.

Usage

When to use it

  • End-to-end lead time dominates the problem: the flow crosses several roles or units.
  • Existing and repeatable process: improving a flow whose current state can be observed or estimated on real cases.
  • Team must agree on the waste: building a visual language common to business and technical teams before designing the solution.
  • Current and future states need comparing: quantifying the gap between the process as it runs and its leaner target to ground an initiative.
  • Development flow needs shortening: mapping the path from the discovery of a solution to its release into production.

When not to use it

  • Non-repeatable or knowledge-intensive work (research, creative work, decisions): durations vary from one case to the next; prefer a qualitative root cause analysis.
  • A single step at fault: a map of the whole flow costs too much; target the step with a root cause analysis.
  • Goal is continuous improvement in small steps: VSM redesigns the flow in one-off overhauls; prefer flow metrics and a Kanban board.

Description

Like SIPOC, which gives a one-page overview of a process, value stream mapping is one of the methods of process analysis. BABOK lists it among the lean methods. Lean is an approach to organising work that grew out of the Toyota Production System: it removes from the flow whatever adds no value for the customer. In BABOK, the map follows the flow from the upstream end of the supply chain to the customer and sets wait time against processing time at each step. The Agile Extension applies it to solutions: there the value stream is the flow of material and information required to bring a solution to the customer.

The value stream and the two maps

Mapping proceeds in two stages whose order is fixed. The team first draws the current-state map, which shows the flow as it works: a team can only improve what it has seen. It then derives the future-state map from it: the same flow rid of avoidable waste. The current map is the object of analysis. The future map is the result of design. A team that draws the target before it has recorded reality defeats the purpose of the exercise.

Drawing the current state

Preparation brings together a cross-functional team: the business experts who work in the flow and the technical roles it draws on (development, testing, operations, architecture, supplier). A business analyst usually facilitates the workshop. The team appoints a value-stream owner who knows the current process. It chooses one product, one product family or one service and sets the scope of the map. Finally it names the value for the customer that the flow must produce. A map holds one family on one page. Beyond that it becomes unreadable.

The team then traces the flow backwards from the customer. It observes or simulates the real path. The process boxes line up from left to right. Under each one, a data box carries its measures: processing time, changeover time, availability rate, number of operators. The information flow runs across the top of the map. It shows what triggers each step: the orders and the schedules. Between the boxes, inventory or queue triangles hold the work that waits. In a service flow, each triangle stands for the requests waiting in a mail tray or an electronic queue.

The team then fills in each data box. It measures processing time or estimates it. An estimate beats an empty box: the more precise the data, the easier the improvement is to target. At each queue, the team records the wait time. The standard notation for these objects comes from Rother and Shook's Learning to See: icons for process, inventory and push flow. In a push flow, each step sends its work to the next one without being asked for it. A map that everyone can read depends on this notation as much as on its figures.

The timeline and cycle efficiency

The map distinguishes three durations. Processing time is the time spent working at a step. The lean tradition also calls it cycle time. Wait time is the time spent in a queue before a step. Total lead time adds the two across the whole flow. The customer experiences it from request to delivery. A team that confuses the cycle time of a step with the total lead time of the flow distorts the whole analysis.

The timeline runs along the bottom of the map in a sawtooth: the upper segments carry the wait times and the lower segments carry the processing times. The team sums them separately. Cycle efficiency relates the time in which the work adds value for the customer to the total lead time. The map records only processing time. Processing time also includes checks and the other steps that bring the customer no value. The ratio calculated from it is therefore a ceiling on cycle efficiency. A low ratio points to the queues as the target of improvement. This timeline separates VSM from a plain flowchart, which shows the sequence without saying how long the request waits.

From current state to future state

The analysis of the current state separates the steps that add value from the rest. It then splits the rest in two. A regulatory check or a statutory deadline imposes necessary waste: the team cannot remove it. The team can eliminate unnecessary waste (excess paperwork, over-long queues). The Agile Extension grounds this analysis in root cause analysis: the team looks for the reason a queue forms before it removes the queue. The team then marks the areas to address on the current map and sets an improvement measure for each one. In the notation of Rother and Shook, the kaizen burst icon flags each of these areas. Finally the team draws the future-state map without the unnecessary waste.

The team then plans the improvement and lists what supports it: information systems, training, changeovers. In an agile setting, changes to the supporting technology become backlog items that feed an agile initiative. Once implemented, the future state becomes the new current state. The next round of mapping can then start from it. The Agile Extension also maps the development value stream from business analysis to release by way of planning, design, build and test. Each step there carries its wait time and its processing time.

During execution, analysis paralysis means perfecting the current-state map instead of moving on to improvement. The team therefore time-boxes the work on the current state. Overload sets in when the team tries to capture everything: the map turns into an unreadable inventory. The scope set during preparation prevents it.

AI considerations

A language model prepares the workshop. It extracts the steps, handovers and queues that a written procedure, interview notes or emails mention. The team then corrects this first draft on the map. The measured durations come from another tool. Where each step leaves a timestamp in a case-management system, a process mining tool computes the observed wait and processing times of each queue from the event logs. Participants underestimate waiting when they estimate it from memory. Measurement corrects that underestimate.

Deciding what counts as waste stays a human judgement. A business expert decides whether a wait is necessary or unnecessary: the expert knows whether a regulatory check or a statutory deadline imposes it. The map must reflect the flow observed in the field. An event log records only what passes through the system: paper mail, emails, phone calls and workarounds escape it. A model fed on logs alone reproduces these blind spots. The team itself agrees on the waste to eliminate: no model makes that decision for it.

Examples

The example follows a common service flow in Switzerland: a health insurer reimburses an outpatient treatment invoice received by post. The durations are fictitious.

Value stream map of the current state. An insured person or a provider sends a paper invoice. Five processing steps follow one another, each preceded by a queue: scanning, completeness check, benefits check, validation, payment and statement. The benefits-management system drives the information flow across the top of the map. A sawtooth timeline sets 20 working hours of waiting against 23 minutes of processing. Cycle efficiency reaches at most 1.9 percent. Kaizen bursts mark the three queues that the future state removes.statementI3 hI5 hI6 hI3 hI3 hScanningthe mailProcessing time4 minCompletenesscheckProcessing time6 minBenefitscheckProcessing time8 minValidationProcessing time3 minPayment andstatementProcessing time2 minInsured / provider(paper invoice)Reimbursed insuredBenefits-managementsystemSawtooth timeline (working hours, 8 h per day)3 h4 min5 h6 min6 h8 min3 h3 min3 h2 minWait20 hProcessing23 minTotal lead time20 h 23 minCycle efficiency1.9%Processing stepQueuePush flowElectronic informationManual informationKaizen burst
Current-state VSM of a health insurer reimbursing an outpatient treatment invoice. The five queues (upper segments of the timeline) total 20 working hours. The five processing steps (lower segments) total 23 minutes: cycle efficiency reaches at most 1.9%. Kaizen bursts mark the three queues that the future state removes.

For the insured person, a total lead time of 20 h 23 min in working hours comes to about two and a half working days. Processing takes only 23 minutes of it: the invoice spends about 98% of the lead time waiting. Each queue costs hours and each step costs minutes. The improvement therefore targets the queues.

A future state sketched on the same map targets the three queues marked with a kaizen burst. The insurer receives invoices electronically and drops scanning along with its three-hour wait. It merges the completeness check and the benefits check into a single verification. The six-hour queue that separated them disappears. The single verification still takes the 14 minutes of work that the two checks took. Finally the insurer validates invoices below a threshold amount with no wait. For these invoices, the lead time falls to 8 h 19 min in working hours, that is about one working day.

Durations in working hours (8 h per day). The future state applies to invoices below the validation threshold.
StepCurrent stateFuture state
Upstream waitProcessingUpstream waitProcessing
Scanning the mail3 h4 minRemoved (electronic invoice)
Completeness check5 h6 min5 h14 min (single verification)
Benefits check6 h8 minMerged with the completeness check
Validation3 h3 min0 h3 min
Payment and statement3 h2 min3 h2 min
Total20 h23 min8 h19 min
Total lead time20 h 23 min8 h 19 min
Cycle efficiency (ceiling)1.9%3.8%

Cost

PhaseLevelJustification
PreparationMediumBringing several functions together at the same time, setting the scope and recording the real flow in the field rather than from memory.
ExecutionLow to mediumOne workshop is enough for a current-state map. The cost rises with the number of families and when the team measures the times instead of estimating them.
DocumentationLowThe deliverable fits on one or two pages: the current map and the future map.

Tools

Teams first draw a value stream map by hand. In an agile setting, a whiteboard and sticky notes are enough for the workshop: the team corrects the map during the session. A spreadsheet then holds the data boxes and computes the sums and cycle efficiency with no other tool. For a wide flow or a map to maintain over time, the team moves to diagramming software such as Lucidchart, draw.io or Visio. Visio offers a value stream map template in its Professional and Premium editions. The eVSM tool is built for value stream mapping alone. A remote workshop runs on an online whiteboard such as Miro or Mural. Where the information system timestamps each step, a process mining tool supplies the measured durations.

Sources

  • IIBA, A Guide to the Business Analysis Body of Knowledge (BABOK Guide) v3, §10.34 Process Analysis: the definition of the technique, its place among the process analysis methods and the limits the guide states for process analysis as a whole.
  • IIBA, Agile Extension to the BABOK Guide v2, §7.23 Value Stream Mapping: the value stream as the flow of material and information required to bring a solution to the customer, the current and future states, the development value stream as an object of mapping and the conversion of the changes that the future state requires into backlog items.
  • Mike Rother and John Shook, Learning to See: Value Stream Mapping to Add Value and Eliminate Muda, Lean Enterprise Institute: the current-state and future-state method and the icons of the map (process, inventory, push flow, kaizen burst).
  • James P. Womack and Daniel T. Jones, Lean Thinking: Banish Waste and Create Wealth in Your Corporation, Free Press, 2003 (revised edition): the value stream as the second principle of lean and as the foundation of mapping.
  • Lean Enterprise Institute, Value-Stream Mapping, Lean Lexicon: the definition of the technique, its origin at Toyota, the current and future states and the measures of the data box.
Value Modelling
All techniques
Vendor Assessment