Too Much Transparency

We are all naturally in favour of greater transparency. We think that if everyone could see the progress of production and levels of stock in the warehouse, as well as shipments and stocks in every warehouse along the supply chain, we should all be better off. We should be able to adjust our actions to changes elsewhere in real time. Add to this the Finance Director’s dream of being able to control everything from a central point and you have the promise of RFID – radio tags on every product telling you where they are at any time. 

But this dream could so easily turn into a nightmare. Just think what happens now. You have plenty of stock in your finished goods warehouse and yet you are always short of the one product the customer desperately needs. So you get on the phone to get production to change their plan and make some of this product in a hurry. Production time and efficiency is lost because you change the plan. People have to scurry round to chase materials and the production of other products is delayed. As a result you are short of those products and the cycle begins all over again. 

In this situation it is easy to blame fickle customers for changing their minds. However there is a more insidious form of variation in orders that is encouraged by greater transparency. If your planners have visibility of all the stock in your finished goods warehouse what is their natural reaction to a run of demand for one product? They change their plan to try to restore stock levels, even if it does not breach the lower limit for stock of this product. 

There is a natural human desire to try to adjust things to the average, even though the point of the warehouse is to absorb variations in demand so these waves do not flow upstream. Why do they do this? Because experience tells them that they never know exactly when production will make another batch of this product – it may not happen for another month or more if the plan gets changed and the next production run gets rescheduled, as it almost certainly will. So better change the plan now rather than wait until you are out of stock. 

Transparency will not actually solve this situation, and could make it a lot worse. Part of the answer is to set the upper and lower stock levels to absorb the variation in demand and stick to them, so these fluctuations are not passed upstream. The other answer is to move away from batch thinking in planning and production and begin to relate the rhythm of production more closely to demand, moving from producing every product once a month to twice a month, and then to every week and maybe twice a week and eventually every day. 

Levelled orders creates the stability that is necessary to start the traditional lean journey in production – improving the capability and availability of each step so you can link them to create flow and then accelerate the rate of flow by improving the frequency. The end result is that production can actually make every product frequently and reliably.

But this is not the end of the story. Sorting out production alone is not enough – you must also sort out your planning logic and behaviour. Just as you eliminate big batches in production you must no longer batch information – it needs to be passed on without manipulation and released to production in small batches frequently. This probably means no longer using your MRP system for production and shipping instructions.

But it probably also means not letting your planners see the stock levels in the inished goods warehouse! We recently witnessed just such a situation – where lanners were only alerted when stock levels breached the upper or lower limit. Even then they learnt not to react too quickly, as they knew that spikes and troughs in demand have a habit of going into reverse very quickly and production is anyhow going to replenish that product in the next period. Doing nothing proved to be the best course of action and eliminated most of the plan changes. 

Imagine this story repeated at many points up the supply chain. Rapid replenishment turns out to be much more important in improving supply chain efficiency than transparency. 

Yours sincerely

Professor Daniel T Jones

Leaning Healthcare

Healthcare is the next great industry to begin the lean journey. The existing model in which the hospital doctor acting as a skilled craftsperson effectively manages their own waiting list of patients, clinics and operations inside someone else’s mass production general hospital is reaching the end of the road. We need to create a vision of what it means to be a lean doctor, what is involved in running a lean health delivery organisation and how the context needs to change to help bring this about.

Recent experience in the UK has shown the problem can not be resolved by spending more money or by increasing capacity and staff. Better outcomes for patients, more satisfying working conditions for staff and lower costs to the tax payer (or members of private healthcare schemes) can only come from fundamentally redesigning the underlying processes for delivering healthcare.

The most important difference between healthcare (and many services) and manufacturing is that the patient is present throughout most of the process, indeed the patient is the product and their problem is the purpose of the activity. If you are a manufacturer just think how different your life would be if your product could experience your process and tell you what it was like!

So healthcare is actually two parallel processes that have to be synchronised: the patient process (which begins and ends at home), the diagnostic and treatment process which mirrors it (in GPs surgeries and hospitals) and several enabling support processes like radiology, pathology, pharmacy, supplies, bed management and theatres.

Healthcare has traditionally focused on the patient doctor interaction and ignored the rest of the patient journey - on waiting lists, searching for a place to park, sitting in queues etc. The introduction of patient choice in the UK is beginning to focus attention on reducing these non value creating steps. Our lean experience tells us that these are also symptoms of lots of wasted time and effort in the diagnostic and treatment processes and in the support processes.

The second characteristic is the huge variety of patients with different conditions coming into the surgery or the hospital. To make sense of this we need to see the different product flows through the healthcare system and begin to manage them separately. What turns out to be critical in defining these flows through a hospital is the length of stay (or the rhythm or takt time in lean language) – whether patients go home that day, stay for a day or two, stay for more than a few days, or whether they need long term care – and then whether they need surgery or other specialist treatments or not (what process routes they follow).

Like manufacturing there is a common assumption that demand is volatile and unpredictable. Experiments with open access to GP surgeries and analysis of people coming into Accident and Emergency Departments shows that demand is actually quite stable and predictable. The greatest variation is in elective work that has been sitting in waiting lists and scheduled and rescheduled many times. Queues (just like inventories) and the scheduling and planning that goes with them actually create significant and unnecessary extra costs throughout the system. The underlying pattern of demand for elective work is also relatively stable.

Having defined the flows (value streams) there is still a strong belief that every patient is different – and cannot be treated like cars going down a production line. However if we sieve the types of problems being treated we quickly see that in each value stream a few problems account for the majority of the work.

Once we create a regular flow of patients with these common problems we can actually free up more time for treating the patients with more unusual problems. Indeed because we have a more predictable process we are better able to tell patients what to expect, and even involve them in managing it.

To create steady flow means starting at the end of the value stream – with discharge! If you are not discharging patients as fast as patients are arriving then the process inevitably gums up. So discharge has to pull patients into beds and through theatres and through admission. This means much greater cooperation between departments, more standard procedures, synchronised test cycles and ward rounds and much clearer and unambiguous handoffs. This is where the lean foundations such as standard work, 5S and problem solving can initially help to improve quality and later as activities are linked to increase the number of patients that can flow through the system. 

We are still at the beginning of the lean journey in healthcare, as courageous pioneers figure out how to do all this in practice. Once we have a better understanding of how lean can transform existing healthcare delivery organisations it will be time to look beyond at innovative new ways of delivering care and at the design of right-sized tools to facilitate them.

In the end healthcare and manufacturing are not so very different. The language and the sequence of changes may differ, but the lean principles work everywhere. Some senior clinicians and chief executives have recently said that “lean can save healthcare”. Manufacturing and service firms and lean experts can help this cause by sharing their knowledge and their experience of lean with local healthcare organisations.

Yours sincerely

Professor Daniel T Jones

Four Curses

I breathed a huge sigh of relief the other day. We found a bookkeeper to manage the accounts for our rapidly growing business! Lizzie, my Operations Director, and I concluded that we were temperamentally unsuited to the task. Now she can spend all her time managing our busy schedules, workshops and conferences. I can do more useful things between assignments. 

It is amazing how we let our lives get sidetracked from the things we know we ought to be doing, by numbers that tell us to do the opposite. Yet we often feel powerless to change the root cause of this conflict so we can get on doing the right things. Here is my initial list of common curses – I am sure you can think of more.

First come forecasts – which are always wrong, including my own! In this day of electronic point of sale and the ability to transmit real sales data upstream almost instantly, why are we still using forecasts? I so often find that delays in the information flows back upstream are longer than the physical flows. If only we could see how information gets hopelessly distorted the more hands it passes through and the older it gets. Would we act this way if information began to rot like a dead fish?

The answer is not as simple as saying we are only going to react to sales data as they happen. We need to take account of changes in trends and to cover a degree of anticipated fluctuation in real demand. In most cases real end-customer demand is actually very stable and the degree of variation for our high volume products and services is actually quite small. So we ought to at least be able to modify our production plans based on rapid feedback from real sales data as they happen. Our model ought to be picking exact quantities of fresh vegetables in the field today for sale in the store tomorrow, based on the weekly pattern of sales and today’s sales data. 

Second is the curse of Economic Order Quantities. This algorithm is the second source of noise in our information flows. Yet the more we learn about managing lean supply chains the more we see that big batches create extra costs elsewhere in the value stream and lead to chronic instability – which in turn causes fire-fighting, expediting and chasing. The world of perfect information and everything always going according to plan simply does not exist. Batching activities separately across a value stream simply makes things worse as their impact is amplified.

Aligning batch sizes for volume products to daily or at least weekly demand makes flow possible and creates the stability for ongoing continuous improvement. Make low volume products to order separately. The same is true in distribution – picking up and delivering small batches of more products from more suppliers on milk rounds turn out to be cheaper than waiting for each supplier to fill a truck with their own products. How much noise do your systems create for your suppliers?

This leads us to the third curse, chimney costing within a department or function. This assumes that by keeping every activity busy by ensuring there is always work waiting to be done, we optimise the utilisation of every asset, department, piece of equipment or truck.

Simply calculating the OEE (Original Equipment Effectiveness) reveals that it actually results in much poorer utilisation. By concentrating on all the causes of interruption in order to synchronise one step with another and create a flow, we end up with much higher utilisation. We are also beginning to optimise the end-to-end flow of value creation and not optimising each activity in isolation. 

This leads to the fourth curse, standard costing. Just looking at the costs of direct labour and slow freight has led to many mistaken location decisions and much longer supply chains than we needed. If we look at all the costs associated with the end-to-end value stream, including all the costs of managing variation, we would take very different decisions and have much more effective supply chains. And we would not be wasting so much of the world’s energy resources and causing so much pollution! 

Challenging these familiar but mistaken rules of thumb will help us do the right things for our customers, our supply chains and our organisations. 

Yours sincerely

Professor Daniel T Jones

Developing Problem Solvers

We traditionally see an organisation as a collection of departments or activities, each managed separately and each separated from the rest by inventories or time buffers between them. Performance is improved by setting targets and budgets. When these are not met we change the managers and if that does not work we restructure the organisation. We instinctively reach for structural solutions because they are quick and relatively easy. However the underlying processes and cost structures remain more or less unchanged.

When I walk round any organisation I see it as a collection of customer processes (if it is a service delivery organisation like a hospital), design and production processes (creating the value the customer is paying for) and many support processes that enable these value creating processes to flow. The task is to identify the value in each of these processes, to see and manage the end-to-end flows and to synchronise the support flows. 

If I can not see the end-to-end flow through production, then neither can employees and managers. So the first task is to help them see their processes and to uncover the reasons why they do not flow. Quite often this means looking at the impossibly complex mix of products they are attempting to flow through their processes. It also means challenging the batch logic of their planning systems trying to schedule every product or batch through every operation. Getting over this hurdle creates the conditions where we can begin to flow most products through the entire process. It also creates the stability necessary to develop standard operations in every process step, which is the baseline for continuous improvement.

As well as looking down at individual processes, I also want to fly a little higher and look down at the organisation as a whole. What are the major flows through the organisation and how do all these processes interrelate? I have in mind a fishbone diagram, overlaid on the organisation chart. The value creating processes form the backbone and all the support processes are the fins. Once we can see how an organisation flows then I am sure we will see even more opportunities for improving it. 

The distinctive thing about lean thinking is that it derives from observing best practice organisations and not from theory (which is why academics have such a hard time understanding lean). The lean principles distil the cumulative experience of thousands of people who have spent their working lives solving the problems that enable processes to flow, and to do so in line with customer demand.  

The core expertise required to create and improve processes is a scientific approach to problem solving close to its source. Every problem is an opportunity to improve the process and every problem is also an opportunity to develop your people. The two go hand in hand. 

So the second thing I look for is how good the organisation is at seeing and surfacing all the interruptions and hiccups in their processes. Are these recorded as they occur and what are the processes for responding to them? Are they delegated to an expert group to solve or is everyone involved in some kind of problem solving activity? 

If so, is there a common approach to problem solving across the organisation and a common language for communicating the diagnosis and the results? Is there a policy deployment framework for aligning and prioritising problem solving activities in line with the business goals of the organisation?

More than anything else do managers lead by developing the abilities of their staff to solve problems, at every level in the organisation and throughout their career? Do employees look up to their superiors for the answers to problems or do managers guide their staff to find the right solution by asking the right questions? 

Answers to these questions reveal the real management challenge and opportunity from lean thinking. Process thinking is fundamental to delivering increased value to customers at lower cost. But this in turn relies on an infrastructure for communication and problem solving and a management committed to continually developing the problem solving capabilities of its people, from the top to the bottom.

Yours sincerely

Professor Daniel T Jones

No More Lean Excuses

I have recently been getting a striking reaction from many senior management audiences. They all agree that products have got vastly better over recent years, but they equally agree that the process of ordering and buying them and getting them serviced has got worse!

They spontaneously tell me that while they love driving their upmarket German premium brand cars, they will not buy another one because of the terrible experiences trying to get them fixed when they go wrong, which they seem to do all too often. Many of them wonder whether they would have the same experience owning a Lexus, built with Toyota’s fabled quality levels. 

This is in fact a symptom of a deeper problem, from which we can learn a lot. A decade ago as the Japanese car makers began to make serious inroads into world markets, European car makers decided their future lay in coming up with more distinctive designs, adding lots of new technologies, offering a greater range of models and a huge choice of specifications so customers could customise their cars to make them distinctive. Business writers called this mass customisation. For a decade or more it seemed to work. Now the chickens are coming home to roost!

The problem is not with the distinctive designs: when they work they really attract attention, as they do unfortunately when they fail to hit the mark! Adding technology is also not the problem, provided it is thoroughly proven to work flawlessly in the highly demanding environment in which we drive our cars and provided it delivers real value to customers, rather than bewildering complexity. The problem is that we can only really cope with so much choice. As our lives get busier we have less time to deal with all these choices. Beyond a certain point the distinctions between different models and options add little or no value at all.

The real problem is when this complexity begins to negatively impinge on the ownership experience and on the cost base and profits of the producer. It is not difficult to add several thousand pounds of options as you specify your uniquely customised car. However when you come to sell the car, maybe even to the same sales person, ou discover that these options make no difference to the trade in price you are ffered! This leaves a nasty taste in the mouth and makes you very reluctant to epeat the experience. Next time you will probably choose a brand that offers a package that fits your needs. Although the dealer loves the extra profit on the options, these cars often sit around for much longer waiting for the next owner, tying up capital in the meantime.

However the crunch comes when dealers are expected to fix all the problems on this bewildering array of products and options, fitted with unreliable technologies. There is no way they can diagnose and solve all these problems and order and stock all the required replacement parts. Not surprisingly their ability to complete service and repair jobs right first time on time has fallen from the typical average around 60% to nearer 30%. That means two thirds of customers are frustrated and disappointed! And they tell their friends and relatives.

Extended warranties mean that manufacturers are paying for all this. Almost certainly most of the profits they traditionally make on selling expensive options and spare parts are now being lost in ballooning warranty costs. Product proliferation is not the way to compete with the new industry leader, Toyota.

The longer term solution is to go back to the lean basics, to embed quality at source into every step in the design and production process and to review the product range and bundle the options. The short term solution is to improve the parts system and dealers’ ability to fix today’s problems. In Lean Solutions we show how the same lean techniques we use in the factory can dramatically improve car dealers’ ability to get jobs fixed right first time on time. Toyota is now beginning to roll these techniques across its dealer networks across the world. It is surprising how other manufacturers are blind to improving this most frequent point of contact with their customers. Their future will critically depend on how long they take to follow Toyota’s example. Most of them still do not measure the right first time on time experience of their customers.

Yours sincerely

Professor Daniel T Jones

Breaking Through to Flow

The other day I began a speech to a leading supplier of medical devices by congratulating them on the absolutely level demand for their products from existing users and on the steady growth in the number of users. The audience laughed when I asked whether this meant they could plan ahead and never needed to change their plans at the last moment! Like most companies they change their plans every day, and sometimes several times a day.

They nodded in agreement that they and not their customers are responsible for the chaos these short term plan changes cause throughout their extended production and distribution system. The good news is that they ought to be able to do something about it themselves. The bad news is that if their customers knew how much extra cost this chaos causes they would be very unwilling to pay for it! The truth is that they, and many others, are still struggling to understand and deal with the underlying causes of the chaos they are dealing with. It is in fact an obstacle to their taking action to go lean.

The chaos actually begins at the customer interface. In this case the product is part of a diagnosis and monitoring process to manage a medical chronic condition. Mapping the consumption process to obtain repeat supplies will reveal opportunities for saving wasted time for the patient and for the doctor.

Mapping the way the product is ordered and delivered will reveal further opportunities for cost savings and for improving the prospect of your product being chosen rather than those of your competitors. Frequent replenishment will reduce inventories and improve availability while at the same time smoothing order signals. This analysis is too important be delegated to a distributor or wholesaler.

This chaos is then passed up the value stream towards production. In this case it takes over 200 days to reach the doctor through several decision points. Why? I am always struck by how little production people know about what happens down the distribution chain. They do not know how long the chain is, what happens closer to customers and how well the overall system fulfils customer demand. The shocking thing is that it is always longer than you imagine and levels of fulfilment from the customer’s perspective are surprisingly low, despite all the inventories in between. This is a sure sign that no one is responsible for redesigning the end-to-end value stream.

But the real culprit that is causing most of the chaos in production is the fact that our planning systems are driven by batch logic (based on economic order quantities) which depends on perfect information. We also believe we must plan every event for every product in the same way. Every time things go wrong we make a new plan and when this does not work we change it again and again. As a result fire-fighting is endemic and production efficiencies are significantly degraded.

A way out of this dilemma is to recognise the damage being done by this batch logic and to learn to see that you can in fact quickly create stability and flow for the few high volume products which account for much of your output. These need to be managed separately from the tail of build-to-order products with low volume and unpredictable demand.

Begin by creating a replenishment pull for these high volume products, absorbing demand variation in a finished goods buffer stock and initially producing them on a fixed volume, fixed sequence cycle. This creates the stability necessary to start down the lean virtuous circle of standard work and continuous improvement. It also allows much faster progress in improving equipment availability, shorter changeover times and integrating production steps into a continuous flow. Over time speed up the cycle, reduce batch sizes and incorporate more products into this flow, and as your capabilities improve vary the volume and the sequence to more closely mirror daily demand.

This path very quickly leads to increased output, near perfect on-time deliveries, much higher employee involvement in continuous improvement and it can be replicated up and down the value stream. You are no longer producing to forecast and no longer need so many planners to rejig the schedule for most of your production. And you will discover that true responsiveness comes from establishing stability and increasing the rhythm throughout the value stream, not from changing plans all the time. Chaos is not inevitable and can be conquered.

We have seen this work in so many different environments where you have a complex mix of products with variable work content or production volumes; from separating different types of service jobs in a car service shop to dealing with different types of insurance claims to separating simple routine from infrequent and difficult operations through hospital theatres. Ian Glenday has now written a workbook to enable everyone to try out this method, called Breaking Through to Flow . This is our first LEA publication and I am confident it will help many of you take the next leap on your lean journey.

Yours sincerely
Professor Daniel T Jones

Lean Lessons for 2006

This is the year when Toyota will almost certainly overtake General Motors to become number one in the global car industry, eclipsing what was once the largest industrial enterprise in the world. Fifteen years ago we predicted this would happen in The Machine that Changed the World. Five years later in Lean Thinking we described what it would take to respond to this challenge. A decade later Western car makers have struggled to build on their first wave of lean improvements. We will also see whether GM and Ford can throw off the shackles of unsustainably high wage rates and huge pension burdens.

The fundamental reason for Toyota’s success is a superior lean business model in which senior managers focus on turning every process into a brilliant process rather than making the numbers and keeping the assets busy in their area. And in which every manager and employee takes responsibility for solving problems to further improve these processes. A problem solving, process focus drives the efforts of the whole company.

Toyota’s triumph will have a huge demonstration effect on every industry across the world. It will accelerate the growing interest in lean, triggered earlier this year by the endorsement from GE that lean is the way forward for them. The auto industry will continue to be a lean reference model for design, production and the coordination of the upstream supply chain. However I expect we will see other industries becoming role models for lean in the near future.

I am now convinced that the consumer goods industry has reached a tipping point. Production lead times are beginning to be dramatically compressed, rapid replenishment pioneered by Tesco is improving availability at lower cost and new retail formats and home delivery are demonstrating that convenience does not need to cost more. The era of the focused factory, big automated warehouses and mega stores is coming to an end. As more and more products, from pharmaceuticals to electrical goods, are sucked down this pipeline this will transform these industries too. Rapid lean replenishment will become a way of life for all.

I am also convinced that leading capital good producers like Rolls Royce aero engines are beginning to see the significant benefits from selling the use of their equipment rather making their money on repair and overhaul. “Power by the hour” transforms the way you design and maintain this kind of equipment – leading to win-win gains for producers and users.

The lean revolution is also beginning to transform service delivery systems from call centres to installation and repair operations. There is a lot we can learn from them about building intelligent feedback loops from customer facing staff and turning every customer interaction into a Kaizen opportunity.

But perhaps the most dramatic lessons will come from applying lean in healthcare. We are beginning to see good examples of how lean can improve the flow of patients through existing hospital processes. This is just the start of a much bigger redesign of complete healthcare systems, combining lean system design with new enabling technologies for diagnosis and treatment. It is here that we will see the fastest progress from process improvement through process redesign to rethinking the whole business model using lean principles.

My hunch is that the discussion at the Frontiers of lean thinking will move sharply towards the customer in 2006. More and more companies will track the frustrations of their customers in accessing and using their products and strive to realise the win-win gains from improving the poor fulfilment of their delivery systems, as we described in Lean Solutions.

A second topic will be how to improve the process for designing new products and the processes to make and deliver them, particularly in the light of lessons from the first round of lean. This includes the whole spectrum from designing complex products with huge teams of engineers to much simpler but smarter ways for smaller firms to introduce new products.

A third topic which lean thinkers are waking up to is what process management really entails, how value stream managers work with functional managers and how to create the problem solving capability in every employee that is bedrock of lean process management. This will be an important topic for research, discussion and experimentation in the lean movement in the year ahead.

I hope you have a good break and return fired up to continue your lean journey. I look forward to meeting many of you again next year.

Yours sincerely
Professor Daniel T Jones