Process Thinking and Value Stream Analysis

Introduction

Lean Management begins with a change in perspective: an organisation should be understood not only through departments, job titles and functional responsibilities, but also through the processes that convert requirements into outcomes. Customers experience processes, not organisational charts. They judge whether a product or service is available, accurate, timely, safe and valuable. Process thinking therefore provides the foundation for identifying waste, improving flow and designing work around customer value.

A process is a connected set of activities that transforms inputs into outputs for a customer or user. The customer may be external, such as a purchaser receiving a product, or internal, such as a planning department receiving verified information from sales. Process analysis makes work visible. It reveals delays, queues, rework, unnecessary movement, duplication, unclear handoffs and constraints that are often hidden when work is viewed only within individual departments.

Value-stream analysis extends process thinking across organisational boundaries. It examines the complete flow of material, information and decisions from an initial request to the delivery of value. This chapter develops practical methods for documenting and evaluating processes, including SIPOC diagrams, flowcharts, spaghetti diagrams, service blueprints, Gemba observation and value-stream mapping. It also explains lead time, cycle time, process efficiency and future-state design.

Learning objectives

By the end of this chapter, the learner should be able to:

• define a process and distinguish value-adding from non-value-adding work;
• construct and interpret a SIPOC diagram;
• use flowcharts, swimlane charts, spaghetti diagrams and service blueprints appropriately;
• conduct a disciplined Gemba observation;
• develop a current-state value-stream map;
• calculate and interpret lead time, cycle time and process efficiency;
• identify waste, bottlenecks, queues and information-flow problems; and
• design a feasible future-state process using Lean principles.

1. Process thinking and customer value

1.1 The process view

A process has five basic characteristics:

1. it has a purpose or customer requirement;
2. it receives one or more inputs;
3. activities transform those inputs;
4. outputs are produced; and
5. performance can be measured.

The transformation may be physical, informational or relational. Manufacturing transforms materials; a hospital transforms patient information and clinical conditions; a university transforms applications and learning activities into educational outcomes; a bank transforms requests and evidence into financial decisions.

A process should be defined at an appropriate level. A process that is too broad becomes difficult to observe, while one that is too narrow hides important interfaces. “Order fulfilment” may be an appropriate level for a value-stream analysis, whereas “enter customer postcode” may be an activity within that process.

Processes commonly cross functional boundaries. A customer order may pass through sales, credit control, planning, procurement, production, dispatch and invoicing. Each department may appear efficient locally while the total process remains slow. Process thinking counters this problem by asking: “How does the complete flow perform for the customer?”

1.2 Value-adding and non-value-adding work

An activity is conventionally classified as value-adding when it changes the product or service in a way the customer requires, is performed correctly the first time and would reasonably be paid for by the customer. Activities that do not satisfy these conditions are non-value-adding from the customer perspective, although some may be necessary under current regulatory or operational conditions.

For example, machining a required feature is value-adding. Waiting for a batch to be released is non-value-adding. A legally required inspection may not directly transform the product, but it can be necessary non-value-adding work. The objective is not to eliminate controls indiscriminately; it is to simplify, mistake-proof or integrate them while preserving safety, legality and quality.

The traditional Lean wastes are defects, overproduction, waiting, non-utilised talent, transportation, inventory, motion and extra-processing. These wastes are symptoms of process conditions such as unstable demand, poor layout, large batches, unclear standards and delayed information.

1.3 Process boundaries and customers

Before mapping a process, the analyst should establish:

• the trigger or starting event;
• the endpoint and deliverable;
• the process customer;
• the unit of work being followed;
• the time period and operating conditions; and
• exclusions and interfaces with neighbouring processes.

A clearly defined unit of analysis is essential. A map that combines urgent and routine orders without distinguishing them may produce misleading averages. Similarly, a hospital process map should identify whether it follows one patient journey, a referral type or all patients in a department.

2. SIPOC: establishing the high-level boundary

SIPOC stands for Suppliers, Inputs, Process, Outputs and Customers. It is a high-level process-definition tool normally prepared before detailed observation. SIPOC prevents premature concentration on individual tasks and helps a team agree on the scope of an improvement project.

A practical method is to begin with the output and customer, then identify the major process steps, inputs and suppliers. The process column generally contains four to seven macro-steps, not every detailed activity. The team should use nouns for inputs and outputs and action verbs for process steps.

SIPOC is not a performance map. It does not normally show sequence, duration, queues or operator movement. Its value lies in establishing a shared boundary and exposing missing stakeholders, inputs or outputs. A SIPOC can also reveal that a process has multiple customers with conflicting requirements, such as speed for the customer and documentation completeness for a regulator.

3. Flowcharts and complementary process maps

3.1 Flowcharts

A flowchart represents the sequence and logic of activities. Common symbols include rectangles for activities, diamonds for decisions, arrows for flow and terminators for start or end points. The level of detail should match the question being investigated. A map intended to analyse approval delays should show decision points and rework loops; a high-level communication map may not need every data-entry action.

A useful flowchart records:

• the activity and responsible role;
• the decision rule;
• inputs and outputs;
• rework or exception paths;
• waiting points; and
• evidence or systems used.

A flowchart should be validated by people who perform the work. Documentation created solely from procedures may describe the intended process rather than the actual process.

3.2 Swimlane diagrams

A swimlane diagram places activities in lanes representing departments, roles, systems or organisations. It is especially useful for identifying handoffs, unclear ownership and repeated movement between functions. Each crossing between lanes is a potential location for delay, information loss or conflicting priorities.

3.3 Spaghetti diagrams

A spaghetti diagram traces the physical movement of a person, material, document or customer on a layout. It is prepared from direct observation, not from assumptions. Excessive crossing, backtracking and repeated visits can indicate poor layout, point-of-use storage problems or fragmented work.

Movement distance can be quantified. If a route consists of segments d₁, d₂, ..., dₙ, then total observed distance is:

Here, dᵢ is the length of segment i and D_total has the same distance unit, such as metres. The calculation is meaningful only when the observation period, route, workload and conditions are specified. A single unusual journey should not be treated as representative without further observation.

3.4 Service blueprints

A service blueprint adapts process mapping to services. It distinguishes customer actions from front-stage employee actions, back-stage activities and support processes. It may also identify the line of interaction, line of visibility and evidence experienced by the customer, such as forms, receipts, screens or waiting areas.

For example, in an online loan application, the customer completes a form; a visible service platform confirms receipt; an invisible credit engine checks data; and a support process retrieves external verification. A failure in a back-stage activity may become visible as a confusing message or unexplained delay. Service blueprints are therefore useful for analysing both operational performance and customer experience.

4. Gemba observation and process evidence

Gemba is a Japanese term commonly translated as “the actual place.” Gemba observation means going to where work occurs to understand the actual process. It is not a surveillance exercise or an opportunity to blame individuals. Its purpose is to compare standard work with actual conditions and to learn from those who perform and receive the process.

A disciplined Gemba observation includes the following steps:

1. define the process, unit and observation question;
2. obtain appropriate permission and explain the purpose;
3. observe the work without prematurely redesigning it;
4. follow an item, customer, document or case end-to-end;
5. record timestamps, queues, defects, interruptions, handoffs and conditions;
6. ask respectful, open questions such as “What happens next?” and “What makes this difficult?”;
7. compare observations across shifts, workers and demand conditions; and
8. validate findings with process participants.

The observer should distinguish fact from interpretation. “Three files were waiting at 10:15” is an observation. “The team is poorly organised” is an unsupported judgement. Photographs, screen recordings or personal data must be collected only with proper authorisation and data-protection controls.

A Gemba checklist may include:

• Is the standard work visible and current?
• Where does work wait?
• What causes the operator to stop or search?
• Which defects or clarifications recur?
• What information is missing at the point of use?
• Is the process safe and ergonomically appropriate?
• Does the actual sequence match the documented sequence?

5. Value-stream mapping

5.1 Purpose and scope

Value-stream mapping (VSM) is a Lean method for representing material and information flow across a value stream. A current-state map shows how work is actually performed; a future-state map shows a proposed pattern of flow and control. VSM is broader than a task flowchart because it links operational activities with demand, inventory or queues, information signals and supplier/customer relationships.

A value stream should normally be selected around a product family, service family or customer journey with sufficiently similar processing steps. Mapping every product and exception at once creates an unreadable map and weakens analysis.

5.2 Current-state mapping procedure

A practical sequence is:

1. identify the customer requirement and product or service family;
2. define the start and end points;
3. collect customer demand, working time and operating schedule;
4. walk the flow from end to beginning and then verify it from beginning to end;
5. record process data at each step;
6. show inventory, queues, information signals and external parties;
7. calculate time-based measures; and
8. identify the major sources of delay, variation and waste.

Typical process data include available time, number of operators, batch size, uptime, changeover time, defect rate, waiting time and cycle time. Data should be labelled with source, date and conditions. A map based on averages can hide a highly variable process, so ranges, medians or distributions may be more informative than a single mean.

5.3 Takt time, cycle time and lead time

Takt time expresses the pace of customer demand. It is calculated as available net production time divided by customer demand during the same period:

If a workday has 420 minutes of net available time and demand is 140 units, takt time is 3 minutes per unit. “Net available time” excludes planned breaks, meetings and other unavailable periods according to the organisation’s convention. Takt time is a demand-based planning rate; it is not the same as the actual processing time.

Cycle time is the elapsed time between completion of consecutive units at a process step, or, in some contexts, the time required to complete one unit at that step. Because organisations use the term differently, the definition must be stated. For a stable single-server process producing one unit every 2.5 minutes, the output interval is 2.5 minutes per unit. A batch process may have a long completion cycle even when individual processing time is short.

Lead time is the elapsed calendar time from the customer request or process start to delivery of the completed output. It includes processing, waiting, transport, queues, approvals and rework. For a serial process, a simplified decomposition is:

The categories must be mutually exclusive for the equation to avoid double counting. If inspection occurs while the item is waiting, it should not be counted in both categories.

5.4 Process efficiency

A common time-based measure is process cycle efficiency (PCE), also called process efficiency in some Lean applications:

Value-adding time and lead time must use the same units and cover the same unit of work. If five minutes of actual value-adding work occurs within a 120-minute lead time, then:

This does not mean that 95.83% of the process can automatically be removed. Some time may be necessary for safety, legal compliance, capacity constraints or customer protection. PCE is a diagnostic indicator, not a complete measure of quality or organisational effectiveness.

Other useful relationships apply under stated conditions. For a stable process, Little’s Law relates average work-in-process (WIP), throughput and flow time:

WIP is the average number of units in the system, throughput rate is units per time period and average flow time is time per unit. The relationship applies to a stable system over a suitably long observation period, where average inflow and outflow are approximately equal. For example, if a process completes 12 cases per hour and the average case remains in the process for 2 hours, average WIP is 24 cases. Little’s Law does not identify the cause of delay; it describes a system relationship.

5.5 Bottlenecks and flow

A bottleneck is a resource or condition that limits the throughput of the system. It may be a machine with insufficient capacity, an approval authority, a specialist skill, a testing stage or a policy constraint. The slowest-looking activity is not always the bottleneck; demand, variability, batch release and interaction with queues must be considered.

Improvement should protect and exploit the constraint, subordinate other work to it, elevate its capacity when justified and reassess the constraint after changes. Increasing upstream production without increasing the bottleneck usually increases WIP and lead time rather than customer output.

6. Future-state design

A future-state map is not a wish list. It is a testable design showing how the process will operate under expected demand and constraints. Common design principles include:

• define value from the customer perspective;
• create continuous flow where product or information can move without waiting;
• use pull rather than pushing work forward solely because capacity is available;
• reduce batch sizes and changeover losses where feasible;
• establish a pacemaker process that sets the rhythm of the value stream;
• control WIP using explicit limits or replenishment rules;
• place quality at the source through prevention and immediate feedback;
• separate standard work from genuinely exceptional work; and
• simplify information flow and make ownership visible.

Future-state design should consider demand variation, staffing, technology, safety, regulation and financial feasibility. It should specify assumptions, implementation actions, measures, owners and review dates. A future-state process may require experiments rather than immediate full-scale deployment.

A useful design sequence is:

1. determine customer demand and takt time;
2. identify the process constraint;
3. remove obvious interruptions and unnecessary handoffs;
4. connect compatible steps into flow;
5. establish pull signals and WIP limits;
6. create standard work and escalation rules;
7. prevent defects at the point of occurrence; and
8. pilot, measure and revise the design.

Technology should support a sound process, not conceal a defective one. Workflow automation can reduce transcription and routing delays, but automating unnecessary approvals may make waste faster rather than eliminate it. Artificial intelligence may assist with classification, demand forecasting or anomaly detection, but its use requires data-quality checks, human accountability, transparency, cybersecurity and controls against discriminatory outcomes.

7. Integrated worked example

An illustrative service process concerns approval of employee equipment requests. A request begins when an employee submits a form and ends when approved equipment is delivered.

Observation produces the following average data for one request:

The value-adding time is defined as 8 + 6 + 12 + 9 = 35 minutes. Therefore:

The map shows that waiting accounts for 285 minutes, or approximately 89.06% of lead time. Interviews reveal that incomplete forms create clarification loops, managers review requests only twice daily, and procurement batches requests at fixed times. A future-state design proposes a standard form with required fields, immediate electronic routing, a daily pull list for procurement and a small controlled queue for urgent requests. The design does not remove approval authority or procurement controls; it reduces avoidable waiting and rework.

The pilot should measure lead time, first-time-right form completion, queue size, approval workload, delivery reliability and user satisfaction. If lead time falls but errors or unauthorised purchases increase, the change has not produced a successful process outcome.

8. Case study: illustrative hospital outpatient referral process

Background

A regional hospital receives referrals from community clinics. Patients report uncertainty about appointment timing, while clinicians report repeated calls to check referral status. Management initially suspects that the scheduling team is understaffed.

Problem definition

The process boundary is set from receipt of a complete referral to the patient’s first appointment. A SIPOC identifies community clinics and electronic systems as suppliers, referral data and clinical information as inputs, and patients, clinicians and the hospital as customers.

Analysis

A swimlane map shows that referrals move between clinical review, administration, scheduling and records. Gemba observation identifies three queues: referrals awaiting completeness checks, referrals awaiting clinical prioritisation and patients awaiting available appointment slots. A spaghetti diagram shows that paper records are transported between two buildings for a portion of the process. The current-state value-stream map records 22 minutes of direct administrative and clinical work but an average lead time of 11 calendar days.

The analysis also reveals that incomplete referrals are returned to clinics without a standard reason code. Staff re-enter information from email attachments into the scheduling system. Consequently, adding staff alone would not address the main causes of delay.

Approach and solution

The future-state design includes:

• a structured electronic referral form with mandatory fields;
• standard reason codes for returned referrals;
• a single digital work queue with visible status;
• defined clinical-priority rules;
• daily review of the constraint queue;
• electronic access to records where legally permitted; and
• a patient notification process that states the current status and next step.

The design preserves clinical judgement and confidentiality. It does not treat an algorithmic priority score as a substitute for professional assessment.

Outcome and learning points

In the illustrative pilot, average lead time falls from 11 days to 6 days, while incomplete referrals decline. The result is attributed not to one technology intervention but to improved input quality, clearer flow, reduced handoffs and active management of the constraint. The hospital continues to monitor safety, equity, patient communication and referral outcomes because speed alone is not an adequate measure of healthcare quality.

The case demonstrates five general lessons: define the boundary carefully; observe the actual work; separate processing from waiting; improve the whole flow rather than one department; and protect professional, legal and ethical requirements while removing waste.

9. Challenges, limitations and governance

Process maps are representations, not reality. They can become outdated, oversimplified or politically influenced. Employees may alter behaviour when observed, and averages can conceal peak-period failure. Analysts should therefore triangulate direct observation, system timestamps, interviews, customer feedback and quality records.

Important limitations include:

• Scope bias: excluding suppliers, customers or exceptions can produce a falsely efficient map.
• Measurement bias: inaccurate timestamps or inconsistent definitions make comparisons unreliable.
• Local optimisation: improving one activity may increase total-system delay.
• Demand variation: a process designed for average demand may fail during peaks.
• Human factors: fatigue, skill differences, interruptions and psychological safety affect actual performance.
• Compliance risk: removing a control without risk assessment can create legal or safety exposure.
• Data governance: maps containing customer, patient or employee information must follow privacy, retention and access requirements.

Ethical process improvement treats workers as sources of knowledge rather than causes of waste. Performance data should not be used to punish individuals for system-level constraints without fair investigation. Changes affecting workload, surveillance, automation or job design should be communicated transparently and assessed for unintended consequences.

10. Best-practice checklist

Before approving a process analysis, confirm that:

• the customer, scope and unit of work are explicit;
• the current state was observed at the Gemba;
• process participants validated the map;
• definitions of cycle time, lead time and value-adding time are documented;
• data periods and operating conditions are stated;
• queues, rework, information flow and exceptions are visible;
• the bottleneck has been investigated rather than assumed;
• future-state changes preserve safety, quality, compliance and equity;
• technology is applied only after process logic is understood; and
• measures include customer, quality, delivery, cost and people outcomes.

Chapter Summary

Process thinking views an organisation as a network of connected transformations that deliver value to customers. SIPOC establishes the high-level boundary, while flowcharts and swimlane diagrams reveal sequence, decisions and ownership. Spaghetti diagrams expose physical movement, and service blueprints connect customer experience with front-stage, back-stage and support activities. Gemba observation provides evidence about the actual process and should be conducted respectfully and systematically.

Value-stream mapping integrates material, information and decision flow across a product or service family. Lead time includes processing and delay; cycle time describes the operating interval or completion time according to an explicitly stated definition; and takt time expresses the pace required by demand. Process cycle efficiency compares value-adding time with total lead time, while Little’s Law links WIP, throughput and average flow time under stable conditions. Future-state design uses flow, pull, WIP control, standard work, quality at source and constraint management. Effective improvement balances customer value, operational performance, employee knowledge, technology, safety, ethics and regulatory obligations.

Key Terms

Process: A connected set of activities that transforms inputs into outputs for a customer.

Customer: The recipient or user of a process output, whether internal or external.

Value-adding activity: Work that changes an output in a way the customer requires and accepts.

Non-value-adding activity: Work that consumes resources without creating customer-required value.

SIPOC: A high-level view of Suppliers, Inputs, Process, Outputs and Customers.

Flowchart: A diagram showing process sequence, decisions and connections.

Swimlane diagram: A process map that assigns activities to roles, departments or systems.

Spaghetti diagram: A layout-based trace of movement by people, materials, documents or customers.

Service blueprint: A service map linking customer actions, visible service actions and backstage processes.

Gemba: The actual place where work occurs.

Value-stream mapping: A method for analysing end-to-end material and information flow.

Lead time: Elapsed time from process initiation or request to completed delivery.

Cycle time: Time between completed units or time to complete a unit at a specified process step.

Takt time: Available net time divided by customer demand.

Work-in-process (WIP): Items currently within a process but not yet completed.

Bottleneck: A constraint that limits system throughput.

Process cycle efficiency: Value-adding time divided by lead time, expressed as a percentage.

Future-state map: A proposed design of how the value stream should operate.

Pull: A control approach in which work is authorised by downstream demand or consumption.

Standard work: The current agreed method, sequence and expected timing for performing a task.

Review Questions

Short Answer Questions

1. What is the primary purpose of a SIPOC diagram?
2. Distinguish lead time from cycle time.
3. What type of waste is commonly revealed by a spaghetti diagram?
4. Why should Gemba observations distinguish facts from interpretations?
5. Define process cycle efficiency.

Long Answer Questions

1. Explain how process thinking differs from a purely functional or departmental view of an organisation.
2. Compare flowcharts, swimlane diagrams, service blueprints and value-stream maps, including the situations in which each is most useful.
3. Explain the procedure for constructing a current-state value-stream map.
4. Discuss how takt time, bottlenecks, WIP and pull contribute to future-state design.
5. Evaluate the limitations and ethical risks of using process data and automation in improvement projects.

Analytical / Application Questions

1. A process has 18 minutes of value-adding work and a lead time of 6 hours. Calculate process cycle efficiency, assuming consistent units, and interpret the result.
2. Design a SIPOC and observation plan for a university student-enrolment process. Identify at least three likely sources of waiting or rework.
3. A service process completes 20 cases per day and has an average lead time of 3 days. Using Little’s Law, estimate average WIP and state the assumptions required for the estimate.

Multiple Choice Questions with Answers

Question 1

Which tool is primarily used to establish the high-level boundaries of a process?

A. SIPOC

B. Spaghetti diagram

C. Control chart

D. Fishbone diagram

Correct Answer: A. SIPOC

Question 2

A process has 420 minutes of net available time and customer demand of 140 units per day. What is the takt time?

A. 1 minute per unit

B. 2 minutes per unit

C. 3 minutes per unit

D. 4 minutes per unit

Correct Answer: C. 3 minutes per unit

Question 3

Which activity is most likely to be classified as value-adding?

A. Waiting for a batch release

B. Re-entering incorrect information

C. Moving a document between offices

D. Creating a customer-required product feature

Correct Answer: D. Creating a customer-required product feature

Question 4

What does a spaghetti diagram primarily analyse?

A. Statistical process variation

B. Physical movement and route distance

C. Supplier financial performance

D. Employee competency levels

Correct Answer: B. Physical movement and route distance

Question 5

A process has 10 minutes of value-adding time and a 200-minute lead time. What is its process cycle efficiency?

A. 2%

B. 5%

C. 10%

D. 20%

Correct Answer: B. 5%

Question 6

Which statement best describes Gemba observation?

A. Analysing work only from written procedures

B. Observing work at the place where it occurs

C. Automating every manual activity

D. Estimating performance from customer complaints alone

Correct Answer: B. Observing work at the place where it occurs

Question 7

Under stable conditions, a process has throughput of 8 cases per hour and average flow time of 2 hours. What is average WIP according to Little’s Law?

A. 4 cases

B. 8 cases

C. 10 cases

D. 16 cases

Correct Answer: D. 16 cases

Question 8

What is the main purpose of a future-state value-stream map?

A. To document only historical performance

B. To show a proposed improved pattern of flow and control

C. To replace all regulatory controls

D. To measure employee attendance

Correct Answer: B. To show a proposed improved pattern of flow and control

Question 9

Which statement about lead time is correct?

A. It includes only hands-on processing time

B. It is always equal to takt time

C. It can include processing, waiting, transport and rework

D. It measures only physical movement

Correct Answer: C. It can include processing, waiting, transport and rework

Question 10

Why is a bottleneck important in value-stream analysis?

A. It determines or constrains the potential throughput of the system

B. It is always the activity with the largest number of employees

C. It identifies the process with the lowest purchase price

D. It proves that all other activities are value-adding

Correct Answer: A. It determines or constrains the potential throughput of the system