Erik Gillet, former global head of operational excellence for financial accounting at Credit Suisse, revisits the five principles of lean applied to services and explores whether the traditional lean wastes are relevant in a service environment.

Many try to lift the lean manufacturing implementation model and drop it into their service environment. Lean in services, however, requires a different approach as the conditions are very different compared to manufacturing. How can we get lean to be an even better fit outside of manufacturing, and what needs to be done differently?

Lean manufacturing tools don’t always apply perfectly to services. Taiichi Ohno developed the “lean” philosophy and tools at Toyota. He was against documenting these principles. He feared Toyota would lose its competitive advantage, but he also believed that improvement is never-ending and codifying the principles would cause the tools to become crystallised.

Exhibit 1 – The Seven Types of Waste Applied to Services

Transportation: Any unnecessary movement of materials, products and/or information – Using mail where e-mail could be used, papers circulating for signatures

Inventory: Anything of value which is being held up from moving forward in the process – Holding unnecessary materials in archives/shelves, e-mails waiting to be read

Motion: Any needless movement of people  Picking up a print out, handling paperwork, clearing away files on the desk, more clicks/keystrokes then needed

Waiting: Any delay between one process step and the next – Waiting for decisions/approvals, system downtime

Overproduction: Any output beyond what is needed for immediate use or customer demand – Working on the wrong thing at the wrong time, unnecessary paperwork

Over-processing: Any effort that does not add value from the customer’s perspective – Redundancies (several people doing the same), creating reports that no one uses, over-collection of inputs

Defects: Any activity related to finding and fixing defects – Data entry errors, missing or inaccurate information, lost records, customer complaints

In the late 90s, lean thinking and tools started to spread outside of manufacturing environments, for instance in financial services and healthcare. Some 15 years later, however, lean theory has hardly advanced. On the contrary, we see practitioners sticking to the tools developed at Toyota and rigidly applying them to every process they work with. Not only is this what Ohno was trying to avoid, it is also like pushing a round peg into a square hole.

Services differ from manufactured goods in various ways. They are generally intangible products. Typically, the customer plays an active role in the service delivery process, and has a personal interaction with the service provider. Additionally, production and consumption of services often happens simultaneously. Service processes are different to manufacturing and consequently improvement tools should also be different.

The philosophy behind lean ascends above and beyond the tools that Toyota developed and used. Lean theory should not be a fixed set of tools, but we should use the philosophy to adapt and expand the toolkit, and to advance lean theory itself. New lean ideas and tools are being developed and leveraged in services.


Lean sees resources utilisation for anything else than value creation for the end customer as wasteful. Such wasted effort should be targeted for elimination. The goal is value creation, but as a consequence lean typically delivers increased process speed, increased capacity, reduced cost and improved quality.

Five principles form the foundation of lean thinking: specify value by product/family; identify all value stream for each product/family; make value flow uninterrupted; let the customer pull value from the producer; and pursue perfection. Although these principles appear generic, their original definitions are manufacturing oriented. Below are some comments to help broaden their applicability to both manufacturing and services.


Many lean efforts focus on efficiency, using technology to improve speed and reduce cost. However, the real competitive edge is value creation. Value should be defined from the (end) customer’s perspective. There are three commonly used categories regarding value:

  • Value added (VA) indicates steps the customer sees as essential: (1) they add form or function, (2) the customer cares (“would pay for it”), and (3) they are done right first time and avoid waste;
  • Value enabling (VE) refers to steps that either allow valueadded tasks to be done better and faster, or are needed to meet legal, fiscal or regulatory requirements;
  • Non-value added (NVA), which is anything not included in the two categories above.

To facilitate discussion around value, this equation can be helpful: Customer Value = (Quality x Utility x Availability)/Price. Processes/activities that do not add to these variables are NVA.


A value stream is the set of activities across all parts of the organisation to deliver a product or service. In other words, the end-toend process delivering value to the customer. Value stream mapping is very powerful to visualise both the production/processing flow and the information flow. This helps to identify waste and its sources. It also creates a common understanding for discussing the current processes and how to improve these.


Eliminating waste makes the product or service “flow” to the customer without any interruption, detour or waiting. Typically NVA activities drive over 80% of time spent in services (waiting time being the most common waste), with VA being less than 10%.


The initial thought behind customer pull was to not produce anything unless the customer asks for it. In practice, however, pull typically focuses on applying just-in-time tools (Kanban cards/ systems). Pull could also be interpreted as understanding customer expectations and demand, and organising process to deliver what the customer wants, how the customer wants it, when the customer wants it.


The prior four principles are not stand-alone. They serve to help reach the theoretical concept of perfect value and zero waste, but this actually means that continuous improvement never stops.


Lean fosters a continuous improvement mindset through three core elements: standardised work, sequencing of process activities, and adding value. Their opposites need to be eliminated, in Japanese: mura (unevenness), muri (overburden) and muda (waste). Yet much of the focus in practice and theory has been on eliminating waste.

Identifying waste is probably the most common application of lean. Originally, seven types of waste were used at Toyota, with several acronyms developed for them (like SWIM TOO, TIM WOOD). These are all iterations of the same elements: transportation, inventory, motion, waiting, overproduction, over-processing (also NVA processing), defects (also scrap). The types of waste were designed with manufacturing processes in mind. As illustrated in Exhibit 1, they can be applied, sometimes forcedly, to services too.

Exhibit 2 – Reducing waste in services – breaking organisational silos

One bank realised that its back-office organisation was functionally siloed and that a stronger internal client focus was needed. To overcome the silos, they implemented “Collaboration Agreements (CAs)”.

Senior managers were asked to identify their key internal customers. Consequently these internal counter-parties were approached to jointly develop a Collaboration Agreement. Essentially the deliverable was a formal contract between these internal parties. Both parties had to work through a set of topics together: What are our roles & responsibilities? What are your and our requirements? How do we measure these? How well do we perform today? How often shall we review our performance?

Although the expectation was a formal document signed by a senior manager on both sides, the amazing observation was that the formal contract actually did not matter at all. It was the dialogue that made the teams realise they had an internal client (often information flows back and forth between teams, making then each other’s client as well as supplier). This bank saw impressive results using this tool: it reduced unnecessary work and improved employee morale, and service delivery performance.

The same bank later introduced “QUICK workshops” (an acronym for QUality Improvements, Collaboration and Knowledge sharing). QUICK workshops were seen as the logical next step to CAs. Where CAs improved the hand-offs between two teams, QUICK allowed to review and improve processes across three or more teams.

Senior managers had to identify critical processes with service quality concerns related to process hand-offs. With the help of the senior sponsor, the functions involved were asked to participate in a facilitated workshop to improve the critical process they perform.

The workshop would start with chronologically walking the process together. With simple questions (What do you need? What do you get? What do you need it for? How good is what you get? What do you do with it?) people started to understand the downstream needs and challenges. This reduced all kinds of non-value added activities, without large investments or additional people required.

It’s interesting to note that mura and muri are often causes for downstream waste. The typical example is sales pulling in orders at the quarter end to meet targets. These orders need to be fulfilled to make the financial numbers, but they overburden the production process and result in waste. Similarly, a process with little standardisation (e.g. unclear instructions or manual activities) could overburden downstream processes as well.

Sometimes an eighth waste is added, relating to the under-utilisation of employees’ abilities (resulting in additional acronyms, like DOWNTIME). The idea is that employees should do meaningful work and add value. Toyota did not list this waste, but reportedly has never laid off people following lean improvements; employees were given new roles instead.

In the spring of 2011, Majdi Alhmah, general manager at BFG International China, started a discussion on the social platform LinkedIn, asking: Besides the traditional 8 wastes: Transport, Inventory, Motion, Waiting, Over-Processing, Over-Production, Defects, and Skills… What do you think will be the 9th waste? He received over 1,600 responses and grouped these in 245 categories. The top five were inefficient communication, complexity and bureaucracy, politics, resistance to change, and unproductive meetings. However, many lean purists argued that the seven original types of waste not only cover everything needed in manufacturing, but are all that is needed in services too: anything additional, they argued, would make things more complex. In other words, the seven types of waste are not only mutually exclusive, but also commonly exhaustive.

It can be argued that not all the top five results are new types of waste, but there is an interesting element to the outcome of this discussion. Apparently, there are types of wasteful activities that have evolved over time, while others (like change resistance and politics) are performed deliberately and covertly. They are harder to spot and to address.


Service delivery processes differ from manufacturing processes. This makes waste in services generally less visible. In addition, service providers tend to have unintended service delivery inconsistencies. These inconsistencies are so particular for services that they should be seen as an additional type of waste. They stem from the characteristics of services, but also from the much stronger human element in the way services are organised and delivered.

Differentiation of service delivery is not necessarily bad. Think about an organisation offering “Platinum” service to its best customers. When we see resource utilisation for anything else than customer value creation as a waste, then unintended variation in service delivery must be a waste too. Variation in either the service delivery process or outcome implies differences in the amount of value added in the absence of an intention, which cannot be the most efficient way to deliver the service.

I see three main causes for inconsistent service delivery: lack of standardisation, unnecessary complexity and redundancies, and sub-optimisation.


Service processes are sometimes less rigidly sequenced and less standardised compared to a manufacturing production process. A lot of waste in services is caused through poor (or lack of) standard work. Activities are done inconsistently, in varying sequencing, or employees improvise because there is no process. As a consequence, service quality will show variation.

Most service processes are repetitive in nature and therefore standardisation helps the organisation to deliver well and consistently. Standard processes can allow for customisation too, but some creativity is needed to come up with configurable and reusable solutions. I often use the term “structured flexibility” in this context. A good illustration of this is mobile phone plans. Subscribers choose the plan that best fits their needs, which gives the telecom provider the ability to standardise its offerings and delivery processes.

Several authors argue that lack of standardisation is a source of competitive advantage (for example through customisation). According to this view, standardisation would diminish the ability to absorb the variety of customer demands. However, it should not be perceived as hindering the ability to absorb variety in customer demand, but rather helping to manage it. Moreover, it actually helps to ensure that the organisation can consistently deliver what is expected. Its absence, resulting in customer dissatisfaction or rework, doesn’t add value.


It is not uncommon for large organisations to solve a problem with a quick fix, or adding a system or data warehouse. Besides not addressing the underlying root causes, the repeated use of these shortcuts creates redundancies and unnecessary complexity. Complexity also comes from process hand-offs and controls. In essence, every hand-off and every decision point in a process adds complexity. This means additional time and effort will be needed, creating opportunities for inconsistencies.

Examples of unnecessary complexity and/or redundancies in services are: duplication of IT systems and data warehouses (with decentralised maintenance); having more than one financial ledger; and multiple forms used for the same process. The generic solution to complexity is simplification: removing complexities, for example implementing a single standard or “version of truth” (for data). If complexity cannot be eliminated right away, find the underlying causes and eliminate/improve these.


Managers typically have functional targets. The larger the organisation, the more likely it is that targets are not aligned and that managers sub-optimise. Typical outcomes of this are poor collaboration across functions (silos), lack of knowledge on activities performed, and possible deviation from achieving overall company goals.

This can also result in institutionalised rework, which may be difficult to recognise. I have seen processes where different departments used different checklists at different stages, therefore not capturing rejects early enough and putting in a lot of effort working on the wrong thing. The generic solution for sub-optimisation is a better alignment of metrics. Ideally this is achieved through a rigorous focus on customer needs.


I would like to share some tools that I did not learn as part of my lean training, but that I have either borrowed or developed as part of improvement initiatives.

  • Voice of the customer

It is not uncommon to hear managers state they know (read: assume) what customers want. An even bigger problem that I have observed is that people in backoffice functions don’t even realise they have (internal) customers; not to mention how well they deliver.

A critical element in the six sigma toolkit is capturing the voice of the customer as a starting point for any improvement initiative. Lean practitioners should more explicitly embrace this practice, to better understand value from the customer’s perspective.

  • Task analysis sheets

Waste in services is not always visible or easily identified. Mapping tools (for example value stream mapping, swim-lane process maps, and spaghetti plots) will help, but do not necessarily provide insights in all the activities performed.

When employees support several service processes, creating an overview of performed tasks helps to visualise and discuss the necessity of all tasks performed. A similar approach can be applied to reduce unnecessary documents.

Task analysis sheets and VA/NVA assessments will identify hidden waste.

  • Collaboration agreements

Many large companies suffer from functional silos that hinder process transparency and efficiency. Process tasks are performed, often without knowing the specific requirements from the internal service recipient.

Collaboration agreements are formal documents that specify quality performance standards between functions. The word “collaboration” instead of “service” is chosen intentionally. In many service processes information flows back and forth between teams. This choice of words allows for more open discussion, without one side acting as the customer and presenting demands to the supplier, and doing so not feeling a need to improve their own way of working too.

Collaboration agreements reconnect processes with (and between) internal customers.

  • QUICK workshops

When functional silos exist it is probable that process tasks are performed across functions, but people performing them don’t have visibility on how their work is used in the downstream process.

QUICK workshops (an acronym for QUality Improvements, Collaboration and Knowledge sharing) are essentially an evolved form of kaizens. They are workshops with at least three functions performing subsequent steps in a service delivery process. They are a logical next step from collaboration agreements; instead of bi-lateral discussions on service requirements and performance, the focus is a more expanded process view.

QUICK workshops translate customer requirements through the organisation, clarify internal hand-off requirements and reduce NVA activities.

  • The 3-step simplification principles

Every hand-off and every decision point in a process adds complexity. The more steps and effort in a process, the more opportunities for process breakdowns.

When Apple developed its first iPod, Steve Jobs got personally involved. He wanted to simplify and would apply the same test repeatedly: if he wanted a song or a function, he should be able to get there in three (intuitive) clicks. To apply this design principle t o user interfaces (including websites) helps developers create an intuitive, user-friendly experience. A similar logic, applied to process hand-offs, also improves customers’ experience.

Reducing hand-offs and duplicate controls will result in less mistakes and more speed.


Lean thinking is all about value creation and aiming for perfection. It is not about rigorously applying the manufacturing tools and techniques that were once documented at Toyota. The lean philosophy remains unchanged, but as its application broadens it is only logical that the theory and tools around process improvement evolve accordingly.