Beyond Squeezing Margins

Over the years I have been invited to speak at many supplier days. The best ones share the stories of the lean progress made by the hosts to encourage their suppliers to follow suit, to mutual advantage. However there is always some tension lurking in the background as both sides size up whether this is a genuine step towards win-win cooperation or merely the same old margin squeezing dressed up in a new guise. The truth usually turns out to be a bit of both.

However it is becoming apparent to me that simply squeezing supplier margins is reaching the end of the road. Without any fundamental changes to the way the shared process works, there is a limit to how much margin there is left to squeeze. More seriously, I am now seeing growing evidence that successive raids on margins are now beginning to have a seriously damaging and even perverse effect on supplier performance. Ignore this at your peril.

This was brought home when visiting a supplier making a basket of different components, which they deliver to a customer for assembly into a complete piece of equipment, made in a variety of different configurations. At the moment the market for this product is buoyant and suppliers are having a hard time keeping up. This is a well intentioned supplier, taking action to respond to pressure from its customers to reduce its prices. However these actions have not always had the intended consequences, for them and their customers.

First, they decided to upgrade their machining plant by buying a completely new set of machines. After a struggle to get all the machines installed and running they are still struggling to get this equipment to work more than 30% of the time! Instead of solving their capacity problems, it made them worse than ever. It turns out the machinery was just not capable of running with these products, to the required tolerances and at the planned volumes. Sure the supplier is making some progress in reducing changeover times and breakdowns, but not nearly enough. This is not just a maintenance problem. More worrying, they were not able to specify the design conditions for what Toyota calls basic stability when choosing the equipment in the first place. It turns out their customer is suffering from the same problem with its new equipment, even though they have been going lean for some years. This is a very expensive lesson for all concerned. Not surprisingly it is a common problem that people do not want to talk much about!

Second, this supplier centralised each of its processes in different “focused factories” across Europe. This was supposed to reduce production costs. But instead most of their products now travel through three or more of their plants before being marshalled for delivery to the customer in their central warehouse. Total lead time through this supplier has gone up and not down, and the probability of having the exact basket of parts ready when the customer wants them has fallen. As a result they are constantly chasing “missing parts”, and their overtime and excess freight bill is enormous.

Third, in order to manage this complex routing through their plants they bought an ERP scheduling system. This turned out to be a disaster. They are still struggling to win the un-winnable war between data that is constantly being undermined by the expediters, trying to end-run schedules based on forecasts that always differ from what the customer actually wants. Schedules to their suppliers jump up and down all the time and not surprisingly on time deliveries from them have fallen.

Fourth, like many they have sourced several key components to low cost sub-contractors in China. Quite apart from all the start up and logistics nightmares, prices in China are beginning to rise. Their sub-contractor is more interested in fulfilling booming domestic demand than their orders, which gyrate far more than planned. They will soon have to find another sub-contractor further inland in China or look for a supplier closer to them in say Romania or Turkey.

This is not an isolated case. Indeed by conventional wisdom they did all the right things. However they were undermined by a lack of real knowledge of the equipment needed to produce their products and by focusing on point economies rather than the cost of the product travelling through the entire supply chain. Every supply chain contains many such stories. Rather than walking away, there are several positive steps that should be on the agenda for the next supplier day.

First, establish an expert working group to pool and upgrade the collective knowledge about specifying right-sized equipment that is fully capable and available. This is a strategic foundation underpinning the shared enterprise.

Second, the customer should, like Toyota, begin to assume responsibility for organising inbound logistics, picking up products from suppliers using frequent milk-rounds. This is brings the heartbeat of the supply chain to the supplier’s door.

Third, by creating level schedules for high volume products and separately scheduling capacity for low volume make-to-order products the customer can create the conditions where suppliers can follow suit. This inevitably leads to freeing up previously hidden capacity and achieving and sustaining record levels of output, while also freeing up the time of many of the planners and expediters to focus on improving the process.

Fourth, suppliers and customers should jointly map their value streams back to raw materials, as we described in Seeing the Whole, over time compressing them by regrouping as many value creating steps as possible in one location, either close to the customer or within trucking distance.

The path beyond margin squeezing is no less relenting. But it needs leadership from the customer to reshape its supplier base and to build a deeper relationship with fewer suppliers, each with a deeper knowledge to solve bigger problems with you on a continuing basis?

Yours sincerely

Professor Daniel T Jones

Rapid, Reflexive, Replenishment

Maybe a good New Year’s resolution for us all is to begin to take responsibility for leaning and compressing our entire supply chains. Ultimately the success of every business is determined by the success of the supply chains of which they are a part, just as a supply chain is only as strong as the links in the chain. Whether we like it or not a supply chain, or more accurately an extended value stream for each product family, is a shared process between all the parties, and needs to be managed as such. 

But where does your value stream begin and end? Probably back to the raw material processing for your longest lead item. At the other end I would argue it does not just end with the consumer purchase of your product – but through the life of the product to its replacement or disposal. Even if we just count back from the point of purchase, how long is your supply chain? Longer than the 319 days to make a cola can or double that to make a pharmaceutical pill? This is something every business should know. 

The second fact everyone should know, but few really do, is how well does this value stream really serve the needs of its end customers? The results will probably shock you. Grocery retailers setting up to supply orders placed on the Internet discovered they could only fulfil about two thirds of the items customers actually ordered, even though the availability of individual items was close to 98%. As a consequence many customers are dissatisfied with the substitutions made on their behalf. Think about how often you found the shoe size you wanted in the style you chose – and then remember that at least one third of the shoes in stock, which you did not want, will be remaindered at the end of the selling season. In other words, how do you disappoint your customers? And how much effort and hassle is required on their part to get what they want from you, if it does not go right first time? 

The true performance of your value streams can only be understood by taking a walk. I was recently reminded of the first value stream walks we did with a combined management team from Tesco and its key suppliers back in 1996. We walked the path of several products back from the store through two warehouses to production and packaging. No one had done this before and it opened their eyes and triggered Tesco’s lean journey. 

After a bit more digging, particularly to follow the order through the information processing maze, it became clear that the way to both improve the fulfilment of the shopper’s basket and to cut swathes of inventories and cost from the system as to dramatically compress the value stream through a series of tight, continuous replenishment loops.

Store sales should trigger replenishment of exactly the same quantities from Distribution Centres. Shipments from Distribution Centres should trigger daily pickups of exactly the same quantities from suppliers. These, in turn, should trigger daily production of the exactly the same quantities, and so on back through packaging and the production of ingredients. The ultimate example of a one-touch, flow-through product is soft drinks placed on rolling dollies at the end of the production line, which are wheeled through distribution to become the shelf fixture from which the customer selects the product. The same logic applies to slower selling products (the majority in most supermarkets) but with either an appropriately longer replenishment cycle – every three days or every week – or by more frequent deliveries of mixed-product shipments of the required quantities.

The model for us, and still the most impressive supply chain in the world, is the Toyota aftermarket parts distribution system we described in Lean Thinking. This still sets the global standard for how to run a lean replenishment system, with lean Distribution Centres, milk-run mixed-load deliveries picking up products and cross docks. 

In the early 1990s Toyota built two highly automated warehouses in the USA and Japan and discovered these could never match the efficiency and flexibility of their manual lean warehouses. They also knew from experience that big centralised ERP scheduling systems can never beat a series of simple reflexive pull loops. Economic Order Quantity logic leads to noise and expediting rather than the optimal use of assets. WebVan, the home shopping firm in California, went bankrupt trying the automated route. And Sainsbury in the UK and Coles Myer in Australia are now struggling because they followed the same path. When will we learn? 

While manufacturers can learn a lot about rapid replenishment from retailing this is not the only place to look for inspiration. Earlier this year I visited a plant making contact lenses. They were busy planning an even bigger, faster machine. This sounded like “hurry up to wait” to me. True enough these lenses went through three different warehouses, each containing mountains of stock and no doubt highly automated, before they reached the customer. And demand for contact lenses is by definition very flat! I suggested they make and ship just the right number of lenses directly to each customer’s home or design simple, but less “efficient” machines that could make these lenses in a local dispensary while customers waited. The room went quiet at this point, until someone said “We never thought of that!” 

What would happen if you applied the same logic to your supply chain? How short could it be? What difference would this make to your customers? What would this do to your investment, design and production costs, and the location of your activities? And how would this change your impact on the environment? 

I look forward to hearing both stories of how long your existing value streams are – and given a blank sheet of paper, how short they could become. 

Yours sincerely
Profeesor Daniel T Jones

Planning for Flow

We have recently visited several food and pharmaceutical manufacturers as they get started on their lean journeys. One of the most challenging was a huge plant making the base ingredients for the pills we buy in our local pharmacy or supermarket. It looks like a cross between an oil refinery and a brewery.  

This plant makes hundreds of different products, each of which goes through a dozen or so processing operations, taking about 15 months from start to finish. These base chemicals are then shipped to another plant to be turned into pills, from which they pass through several warehousing steps before reaching the point of sale. I was staggered to discover that it takes almost twice as long to make a pill as the 319 days it took to make the famous soft drink can in Lean Thinking! 

Not only was there no flow, but quite obviously there was also no pull or levelling either. Indeed we were told that production schedules in both plants change all the time and that fire-fighting is endemic. Like many firms they have pursued lots of initiatives and IT fixes to try to improve their forecasts and schedules, but have little to show for them. This suggests to me they are addressing the symptoms and not the root cause of the problem, which lies in the logic on which their planning is based. The second hurdle to overcome is the strong belief that flow, pull and levelling are impossible dreams in a process industry. 

The logic behind the way production is planned today is based on four key assumptions. First that customer demand is erratic and unpredictable. Second that therefore one needs to be “flexible” enough to make any product in any one of the many mixing tanks, fermentation vessels etc. Third that one will get “optimal use” from all these assets by scheduling large batches (or Economic Order Quantities - EOQ) through each operation in turn. And fourth that the data on which these plans are made is accurate enough to be relied upon. On closer inspection none of these assumptions turn out to be true. 

The consumption of pills, and most food products for that matter, is actually quite steady and does not vary a lot. Most of the variation in orders received by plants is “created” by the way information is batched and passed upstream and by the delays in the many “just-in-case” warehouses on the way to the customer. This would suggest that levelling production in line with demand is both desirable and possible. 

If this is the case then why let the tail wag the dog? Why mortgage the performance of the whole plant for the “flexibility” required for a tiny fraction of production? Is it not more important to create a flow for high volume products so they can be produced in line with demand, rather than constantly recalculating the plan for every product based on EOQ? 

It is also common to find that isolated pieces of equipment whose use is “optimised” using plans based on EOQ end up producing saleable product for a much smaller fraction of the time than equipment whose capability and availability has been improved so they can be linked to enable products to flow through them without interruption. 

Finally people are beginning to realise that centralised control systems can never cope with the minute by minute disturbances that happen all the time on the ground. Events change faster than the plan, beginning a vicious circle of bad data feeding bad plans. Radio Frequency tags (RFID) on every product may help, but may also lead to data overload that turns out to be nightmarishly expensive to fix. Most firms do not know the level of data inaccuracy in their systems. In my experience when they take a look they get a nasty shock. This suggests they might be better off by creating rapid, reflexive replenishment loops all the way back from the customer, which pull rather than push products through production and distribution. 

In our experience, once the fragility of these assumptions has been laid on the table it is possible to open people’s minds to creating flow, pull and levelling. The place to start is by using a simple, but powerful tool developed by my colleague Ian Glenday, which we have come to call the “Glenday Sieve”. By ranking the products (SKUs) by volume, it reveals how skewed the distribution is. It is very common to find that five percent of the product lines account for half the output of the plant. At the other end of the scale one third of the product lines account for less than one percent of output. 

In fact almost every plant (hospital, office process, etc.) has at least two different types of flow that need to be planned and managed differently. In one case you know the high volume products you will produce every week, but not the precise volume. In the other case you probably know the overall capacity needed, but not which products will be required that week. Start by separating out the high volume products, before moving on to analyse what needs to be done to the remaining product lines in order for them to be able to join the flow. 

One of the quickest ways to overcome the sceptics and demonstrate that flow and levelling are possible in a plant that has never experienced flow is by dedicating a set of equipment to produce the high volume products in a regular, fixed sequence. At least to start with. This can be done the following week. This will quickly reveal all the problems with the equipment itself and with the lack of standard work. Provided management can keep its nerve as problems crop up and signal their intention to make flow happen, then the sceptics will come on board. The biggest surprise is that within weeks you can make a lot more products through this equipment than ever before. Now you have secured top management support. 

As employees learn through repetition (and incorporate this into standard work), as equipment availability is improved and as change-over times are cut then you can step-by-step reduce batch sizes, add lower volume products, increase the frequency of the cycle and flex the volumes in each cycle in line with demand. Raw material deliveries for these products can be synchronised with production and planning becomes extremely simple, requiring no fire fighting. 

In parallel there is a lot of work to be done to analyse whether the company needs the tail of low volume products and if so how to schedule these – on separate equipment, at a different time or by reserving some capacity for them in a mixed-model pull system. The end objective, which now begins to look possible through it may still be some way off, is to be able to make every product every cycle in line with demand (to Takt time) with levelled production. A realisable dream. 

As lean thinkers we need to pay close attention to both the needs of the type of process we are dealing with, but also to the tactics that build management support to make it happen and the knowledge base that will sustain it over time. Circumstances will differ, and therefore the emphasis and sequence of actions will too, but the underlying logic and objectives are common. 

I wish you a restful break before getting back to your lean journey in the New Year. Thank you for your interest and support in 2004. 

Yours sincerely
Professor Daniel T Jones

Is Lean So Hard?

We recently had a visit from the Managing Director of a small engineering firm seeking help on his lean journey. His first port of call was his local university, one of the most prestigious in the country. They sent two postgraduate students to show him how to draw a Value Stream Map. However they stumbled when it caine to deciding what to do next. I see this all the time - firms with Current State maps but no Future State map and no Action Plan to implement it. Yet this is one of the key steps in going lean.

By chance I met these same postgraduate students at the Manufacturer Live event in Telford in September. They stnick tue as smart students with a good knowledge of the lean tools. Yet they confessed they did not know how to construct a Future State map for this plant, despite having read the right books. This set me thinking. With the plethora of advice on lean and all the workshops on value stream mapping, why is it so hard to make this essential step, even for the smartest people? What is holding you back?

There are of course some obvious reasons. If no one is given responsibility for straightening out the value stream as it crosses departments then no one is really going to bother to draw a Future State map, let alone implement it. If you do not have support from top management then even the best intentioned lean initiatives are going to run into the sand. Also if your first map reveals so much low hanging fruit then it is not surprising if people go after that, rather than do the more heavy lifting in changing the way things work.

Another common reason is that employees recognise the need to move from making large batches to flowing products through the plant. But they are frustrated in doing so by the lack of basic stability in their operations and in their equipment. Which is why so much attention needs to be devoted to creating standard operations, improving machine availability, reducing changeover times, improving bottleneck processes, etc.

But even if this is a long road there is a danger of pursuing this work without a clear plan of march. You can spend a lot of time creating islands of stability that are hard to sustain unless they are tightly linked. A Lean Value Stream Plan from a Future State map is the way to leverage the synergy between Six Sigma, TPM and Lean and to create flow that lasts.

Another stumbling block seems to be the concepts of takt time and the pacemaker process — how to establish the appropriate rhythm for the value stream and where to trigger it. A really helpful insight is to recognise that you are almost certainly making several different types of products with quite different demand characteristics that require different and not common solutions. People often think lean is about building evetything to order — whereas this is not always the case — in manyy cases it is about rapidly replenishing stock the customer has just purchased.

If you start by analysing you product families by process route and by
frequency of demand and you will discover a few high volume products that account for the bulk of your output. These should be made-to-stock with the customer pulling from a pacemaker at the end of the value stream. At the other end of the scale you may well have a tail of low volume products. accounting for a small fraction of your output that have to be made-to-order from a pacemaker at the beginning of the value stream. Map these value streams separately and treat them as two quite separate projects. Over time it may well be possible to combine these two into a mixed-model pull system, but probably not initially.

The final stumbling block is hidden in your information flow. Where sabilitv is the foundation for creating flow, heijunka or levelling is the foundation for creating pull. Without levelling you are fighting an uphill battle against constantly changing schedules and fire-fighting. We have been brainwashed to think that the only way round this problem is by holding stocks and better IT systems that can improve the forecasts on which our schedules are built. In fact there is a lot we can do to smooth the order signal from our customers. However constantly changing schedules are in fact a symptom of a deeper problem - the batch logic in our scheduling systems. I will return to this topic in a future e-letter.

You may well have encountered different obstacles in deciding what your Future State map should look like. Others of you might have created your Future State maps but have struggled to implement them. I would be interested to hear about both problems, and how you think they can be overcome. Difficult questions are rich food for Lean Thinkers to ponder upon. If we can’t crack this one then we arc not going to make much progress with lean.

Yours sincerelv
Professor Daniel T Jones

Beyond Cost Cutting

What is in your plan for going lean over the next year? Do you still think of lean as just a programme for eliminating waste and cutting costs? Have you woken up to the full potential of applying lean thinking to every process in your business? 

Toyota’s latest Annual Report just arrived on my desk. While most Annual Reports are really boring, Toyota’s are in my view essential reading. Every year it spells out with absolute clarity what Toyota intends to do over the next year. And then it goes ahead and does it! It seldom misses a target and is quite frank about the challenges it faces, which it calls its new frontiers. Think of it as a high level summary of their A3 reports on the plans they are working on right now. 

The first new frontier is its investment in new engine, safety and driver assistance technologies. Toyota was always thought of as a conservative follower taking incremental steps forward, rather than as an innovator. However times have changed. In engine technology, it is already way ahead of the competition with its second generation hybrid powered car, the Prius. Hybrids are going to rival diesel engines as the power source of the future, particularly in North America, where they will appeal as the “guilt-free” large car or truck. Toyota began its “out of the box” thinking about green technologies many years ago. It has now shown that its lean product development process can turn these innovations into marketable products faster than anyone else. What are you doing to anticipate, rather than follow the challenges from your marketplace? 

Second on their list is production engineering. Here their slogan is “Lean and Simple”. They are rethinking every one of their production processes, so they can be redesigned to save cost and so they can be used by less experienced employees in their new facilities across the world. This means for instance “reducing die casting, forging and plastic injection moulds to between one third and one tenth of their "original sizes”, and “reducing the length of their new engine lines by between two thirds and fife sixths”. This is also “out of the box” thinking. 

We should not miss the significance of this development. Most engineers designing new equipment (the same applies to new software systems) will dream of the even bigger, better, faster and more capable machine. As a result you often see the ridiculous situation of a huge line of big machines stretching across the shop floor making a piece that can fit into your hand!

I recently visited just such a plant in Germany, where they love bigger machines. Luckily this firm now realises that bigger machines are not the future. Like Toyota they are busy designing simpler, cheaper and more manual systems for their next generation products. They also showed me some prototypes of their next generation modular, desk-top machines that can be combined in any sequence to make a wide range of small products. These machines also require less operator knowledge and can be moved round the world very easily.  

This greatly facilitates the compression of each value stream, so that as many value creating steps as possible can be placed close to each other. Not only are these clever but simple machines less expensive, but capacity can be added (and removed) in smaller increments to mirror changes in demand over the lifetime of the product in each region. 

Perhaps the most interesting thing is that designing smaller, smarter and simpler machines is an even more exciting challenge for the next generation of engineers than designing the next even bigger machine. In my view every business is going to have to think about this in the future, rather than relying on general purpose machines bought off the shelf. Are you working to rethink and simplify the design of your equipment and production systems for the future?

In addition to this “out of the box” thinking Toyota is continuing its aggressive cost reduction programme and increasing its capacity round the world by leaps and bounds, particularly in China. This global expansion presents Toyota with its biggest challenge, developing enough managers and engineers to run all these operations with sufficient knowledge of the Toyota Way. To this end Toyota has established the Toyota Institute to train future leaders and a new Global Production Centre at its Motomachi plant to train managers to run its plants across the world. 

Many of you will have heard that Toyota attributes its success to brilliant processes and to a production system designed on lean principles. However this in turn is underpinned by a deep knowledge base of problem solving and process redesign. This knowledge is built up as every employee goes through successive rounds of problem solving and root cause analysis, which are captured in a common format, an A3 report. It will be a challenge for Toyota to speed up this knowledge acquisition by new employees from new locations without diluting their performance. Following Toyota’s example it is clear that the big gains from lean come from feeding back lean knowledge into the design of the next generation product and equipment – and from thinking “outside the box”. The real question is how much time and effort you are planning to invest in making this next lean leap, beyond today’s cost cutting? 

Jim Womack and I will be picking up several of these themes at the Lean Management Summit in Aachen, Germany on 11-12 November. I wish you the best of luck in thinking through your lean plans for next year. 

Yours sincerely
Professor Daniel T Jones

Eliminating Failure Demand

Lean Thinkers never take the current state for granted. For example, rather than optimising the flow of products through a bottleneck process, they think hard about how to design right-sized equipment to insert in the value stream and eliminate the bottleneck. Similarly, instead of taking gyrating orders from customers for granted and buffering themselves against them, they know from experience that most of this volatility can be eliminated by working with customers to redesign and smooth the order flow. 

The same thinking is needed when leaning office processes – ranging from call centres dealing directly with end customers to administrative processes supporting the shop floor. The starting point is not how to standardise and industrialise current processes. Instead lean thinkers ask questions about the type of demand they are dealing with? In particular, they ask how much time is spent dealing with problems that arise because the process itself is broken. The usual answer: A lot! 

We heard a great example of this at our Lean Service Summit in Amsterdam in June. A division of Fujitsu Services offers customer and technical support services for large organisations. It keeps equipment and systems working by operating help desks for employees and their customers. Many of these customer and technical support activities have been moving offshore to low wage locations. 

Fujitsu was convinced there is another way to think about this. Instead of worrying about reducing the cost per call answered, they analyse the reasons people are calling. A majority of the calls are, what Fujitsu calls, “failure” demand. They are only necessary because the system or the process is broken and fails to deliver. 

To test this simple hypothesis in your own experience, just think about the last time you tried to get help from a help desk – whether internal or external. Why were you ringing? (Surely you had something better to do!) Did they manage to help you in a way that the problem wouldn’t recur? Was this a satisfying experience? For most of us, we did have something better to do, the problem was not permanently solved, and the experience was frustrating. 

Because these experiences are so widespread, Fujitsu set to work to understand the underlying causes of “failure” demand and how to eliminate it. Typically, their experienced people could quickly track the root causes and redesign the system to remove them. And then the calls fell away! 

What Fujitsu also learnt by analysing demand was a whole lot about what their callers were actually trying to do – about their underlying purpose – in using the software or hardware in question. This gave Fujitsu the opportunity to think about new ways in which systems could be developed to provide additional value for callers, a terrific benefit for the organisations they serve.

Other help desk operators have steered clear of this approach in the belief that their business would disappear. But Fujitsu has got plaudits and a growing stream of work by making both the end users of the systems and their designers of the systems happy. It’s a win-win-win. 

These principles apply to whatever processes you are operating. Just take a look at your activities and ask how much time you and your colleagues are spending on “failure” demand and the consequences for your customers. The shop floor got the “right first time on time” message years ago. It is now time for offices to follow suit. 

Once you can see your “actual” demand (total demand less “failure” demand) – and in the process learn who the customers for your office process really are - you are ready to identify the different categories of need and the steps you carry out to meet them. It may make sense to separate different types of needs and handle them separately. For example you may need to separate the routine, easy-to-process tasks from the messy exceptions. It will also be important to look at your activities by frequency and by volume. It will then be apparent which activities are amenable to standardisation.

You are now ready to embark on your lean process redesign, creating your current state map and working step by step to achieve your future state. Office processes are different in several respects, which need to be taken into account when redesigning the value stream: Process steps are not easy to see and no one can see the whole process from start to finish. Work rarely goes as planned and right first time is very low. Total elapsed time is very long and there are many handoffs. It is hard to see the root causes of waste and it is hard to track progress and problems as they happen. But – and this is the most important insight - they are processes just like production – a series of steps that must be performed correctly in the proper sequence to create the right value for the customer every time. 

There are also a lot of misunderstandings about standardising work in the office and fears that standardisation will stifle creativity. (We heard the same laments in the factory years ago, where they proved groundless.) Identifying and standardising the many routine, repeatable steps and the handoffs between them enables the routine process to run smoothly. This frees up time for creativity in responding to individual customer needs or designing new ways to create value for them. This is the real outlet for creativity and leads directly to better employee satisfaction along with more secure jobs.

Every organisation is a collection of processes, serving internal or external customers. Many of these office and service processes have hardly been touched so far. This is one of the great opportunities in the years ahead, and one that will have a significant bearing on future living standards in countries like the UK. 

You can read more about the Fujitsu example on the articles page of our revised web site www.leanuk.org, where you will also find the report on the Lean Service Summit. I shall also be discussing these issues in more detail at the Manufacturer Live event in Telford on 15th September; maybe I will see you there. 

Yours sincerely
Professor Daniel T Jones

Learning to See Demand

Just how old is the information you are acting on today? How many days ago was the order triggered that you are now producing? And how many days ago did the customer ring up about the problem you are now fixing?  I recently come across two striking examples where the delays in passing  the information upstream were substantially longer than the time it took  to make and ship the product or to solve the customers’ problem. As we  streamline our physical processes so we expose the many problems in our  information processes.  

Even with today’s pervasive data processing and communications the  information process is as gummed up as it has ever been. Mapping your  information flow will reveal a spaghetti diagram far worse than the shop  floor! And one thing we know about information is that the older it is the  less useful and even misleading it becomes.

What we do not see as clearly is that these delays cost money and really  frustrate our customers. We (and they) are paying for the extra buffer  stocks and the excess capacity you have to keep, for the time your managers and staff spend expediting and chasing and for the lost sales and discounts to clear unwanted products.

What we also do not recognise is that it obscures our view of true customer demand and is the biggest obstacle to dropping the gains from lean to the bottom line. We kid ourselves that we can make progress with lean while ignoring the need to redesign the information flow from our customers and within our own organisation.

The knee jerk reaction is to buy a new IT system. Many who have been through the pain of installing a new ERP system know this is not necessarily the answer! Lean thinkers start by asking some fundamental questions about the current process before deciding what IT support is needed.

A good starting point is to really understand your demand – to learn to distinguish between “actual” demand, “created” demand and “failure” demand (which we will return to in the future). At each link in the chain towards your final customer you ought to be able to identify what they actually needed to solve their problem or to fulfil orders from their customers. Compare this with the orders they sent upstream to you and you will see big discrepancy – the sign all clearly contains a lot more noise than the pure “actual” demand signal. 

Toyota calls this “created” demand and it has many causes. Some of it might be caused by your customer using up spare cash at the end of a budgeting period and some may be caused by your marketing staff offering discounts to make their numbers. But most of the noise is created by all the batching, delays and handoffs in the information flow between the customers’ point of use and your point of production. This will be amplified by pervasive errors in your data and the just-in-case factors built into the algorithms in your software. And the noise is even worse if your product passes through several warehouses before reaching your customer.

In other words this noise is mostly due to the way the information system works and not because customer use of your product is inherently volatile. In other words it is something we can do something about - we do not have to take chaotic demand signals for granted! 

The lean thinkers answer is not big centralised systems designed to optimise each activity in isolation – production, transport, distribution etc. Instead they separate production and shipping instructions from materials and capacity planning. They cut the number of decision points to one and replenish exactly what was sold frequently and often. This removes most of the signal noise and optimises the flow of the whole system rather than each operation. And it is simple and foolproof and any deviations can be spotted and responded to as they happen. 

This is exactly what Toyota does all the way up their parts distribution system. After many years work Tesco and some of their suppliers are also close to triggering production and shipping in response to continuous, real time data from their sales tills. 

Yours sincerely
Professor Daniel T Jones