Lean, Agile – and Simple
Experiences from the front line of Operations Improvement, by:
Ferdinand Fanshawe, Sales Director and Martha Ross, Operations Director
The Widget Pharma Company, Orangeville
About 3 years ago our company experienced significant market growth due to the success of some new products. Our business plan reflected substantial growth, with the objective of becoming number 1 in our country marketplace by 200x, but we did not know how to plan for such substantial growth. Our Chief Executive was chatting to a neighbour at their annual neighbourhood barbeque, and he suggested we should look at implementing something called Sales & Operations Planning. Apparently, this has been around for many years, but we had never come across it in our neck of the woods. So we contacted our in-house consultants, and they agreed to send across a couple of people to help us get started. There begins a long story. So first let us say a little more about our working environment.
Our Company Environment
Our company was formed as long ago as the late 1800’s, when we started importing pharma products from the USA. We started manufacturing in the 1930’s, with just the one location, but have grown to a second location within the past 18 months, with the acquisition of a local pharmaceutical operation, thus expanding our product range. We make drugs for both the prescription and over the counter (OTC) markets, primarily for local
customers, but with a couple of major export customers on the other side of the globe. So we make to stock for our local in-country distribution centres (DCs), of which there are 8; there is a local warehouse for the factory, from which goods are shipped to the DCs and abroad. We have around 150 presentations, or SKUs (stock keeping units), based upon about 50 different drug preparations, including both tablets and injectibles; some have just 1 presentation, but others have many. (A presentation is the word we use for stocked end items, when all variations of packaging, box quantities and language are taken into account).
We are planning to hit $100 million this year, so our systems need to be robust enough to deliver those kinds of numbers. We have a competent ERP system, that, when we bought it, was one of the market leaders; but we are probably not using it to maximum benefit, as we had to rush its implementation before the millennium bug hit us in 2000.
We implemented S&OP about 2 years ago, but it did not seem to have brought us the benefits and stability we were looking for. So we thought we should get an external consultant to come in, conduct an S&OP workshop, in which we could measure how well we had done, and identify the areas we needed to improve. Our Chief Executive had bought us some S&OP booklets, and the author seemed to have the right credentials for our purpose. Our workshop was a great success, but when the consultant started, we thought he was talking a different language. Where had we gone wrong? Or did we have the wrong ‘expert’?
Where Had We Gone Wrong?
We found that we had not really implemented S&OP at all. We had a pale imitation. Our process was too short term – looking at the next 1 to 2 months – we should have been looking medium to long term, 12 to 18 months, so we could plan the acquisition of long lead time resources, which are certainly more than 1 to 2 months! We were doing it at the SKU (finished product) level, rather than at the product family level – top management do not have the time to look at individual SKU’s, so we should have grouped these into ‘families’ and reviewed the big picture! We were only looking at the ‘A’ class items, which although they represented around 80% of the revenue, covered only 15% of our product range. We had not defined families. The finance department were not involved – S&OP should be about matching the medium term sales and operations plans to the financial numbers in the budget and the business plan. There was no measure of supply and demand variance, neither monthly, nor cumulatively – we needed to use past performance to help us plan the medium term. We had too much of the ‘silo’ mentality, passing numbers over the ‘wall’ from one department to another – S&OP is about agreeing plans across the whole business! We were not doing medium to long term resource planning at all. We realised that we had policies and procedures that were both incomplete and out of date. We had adjusted the process as we went along, but had not updated our quality manuals. We had no clear inventory policies. There were probably more problems, but that exposes the flavour of our errors. We have since discovered other companies who have made similar mistakes.
S&OP – How Did We Survive Without It? 2 of 7
We were pleased, however, to be complimented on some of what we had achieved. We did have a monthly process, which gave us some analysis of what was going on. We did have a pre-S&OP meeting, which did make proposals to the Senior Management Team. There was some inter-departmental discussion and
negotiation. We do have up to 5 years of history for the established products, by SKU, and we had developed a simple forecasting tool to help us forecast the future demand. The New Product Introduction, NPI, process was providing the necessary updates on a quarterly basis. The process leading up to the pre-S&OP meeting was similar to what it should be, but was looking at some SKUs for 1-2 months, rather than all families through a much longer horizon. We had a schedule of dates for various meetings, going forwards 2 years, such that all meeting participants could book and ensure their availability to attend without fail. However, the meeting was not fit for purpose.
So the workshop developed an Action Plan, and the following describes where we are now up to. We are not yet perfect, but we are getting there. We realise that S&OP is about Sales, Operations, Finance and others planning as one. The process, like all other business processes, needs to deliver continuous improvement.
1. Future Product Planning, NPD, NPI
Our New Product Development (NPD) people have always had a quarterly review process that provides, to both sales and operations, an outlook on the introduction of new products. They continue with this in the same manner as before, but they now also perform a monthly mini-review, and produce a summary monthly update of any major delays or advances of their plans. Sales & Marketing need to consider the impact upon their demand plans and selling activities. Operations need to know whether and when any new products will require new manufacturing processes, or new materials, so that they can start to research a relevant source of supply. We needed to bring this planning into an integrated process, in which reconciliation could occur across the business. The timing of this process has also been changed. It is completed by the end of the month, so that the NPI status update can be provided to the sales and marketing organisation at the same time as the sales analysis at the end of the month.
This is a good example of S&OP bringing everything together into a total framework. Of course, we had a new product development process before S&OP. The problem was that it was not well integrated into the big picture, and the horizon was not always planning far enough forwards.
2. Demand Planning
We have moved away from referring to this as forecasting. That is too narrow a term for what we now do. We like the idea of calling it a demand plan. Think about the difference between a forecast and a plan. A forecast is something we can do little to affect – a good example is the weather forecast. A plan is something we aim to execute. Forecasting is but one element of our demand planning process, which we complete in the first 10 days of every month. Demand planning takes the forecast as the starting point, based upon history, and takes into consideration other relevant input, such as market intelligence, to turn the forecast into a demand plan, that sales commits to delivering as customer orders. Each month the sales and marketing organisations update their demand plan, and feed it to operations so that they can see if they can deliver it.
The first thing we changed was our horizons. We now look 18 months forwards, on a rolling basis. This means that when we are half way through the current year, we already have a view of the whole of the following year. This means that our budgeting process for next year starts from the ‘next year plan’, not from a blank sheet of paper, which is how we used to do it. We have the first 12 months in monthly buckets, and the next 6 months in quarters. This gives us the level of detail we need.
As we reach the end of the month, we close the books on the sales orders for last month. We update all the data in the database, all at the SKU level. We now have an extra month of history. We analyse the data in various dimensions for dissemination to the regions. We identify outliers (where the numbers just look plain wrong), and ask the relevant people to advise whether we should accept each outlier ‘as is’, or whether the numbers should be adjusted. Next, we run our forecasting program.
We were using a 3-month moving average, but the workshop helped us understand this was inadequate in a market with clear seasonal demand patterns. We have now incorporated seasonal indices into our forecasting program, and we have great expectations of improved forecasting accuracy. We are now looking at commercial forecasting software, to see if we can improve on what we already do. We are now forecasting all SKUs.
S&OP – How Did We Survive Without It? 3 of 7
But, be aware, the forecast is but one part of our demand plan. We considered using the terms demand plan or sales plan, but settled upon demand plan, which seems to be a more common term. When the forecast has been created, it too is disseminated to all interested parties, who provide input to the demand planning process – the sales regions (liaising with their major customers), marketing, product development, demand planners/managers and the sales director. They all review the numbers, and feed back to the demand managers, who collate the various data inputs to create their draft demand plan for the demand planning meeting. All the data is collected at the SKU level. The forecasting and feedback is at the SKU level. The numbers for the demand planning meeting are shown both at the SKU and at the family level, added up from the SKU level.
We realised we could not conduct the monthly meetings at the SKU level, so we grouped the SKU’s into families. We know that, ideally, the family definitions for both sales and operations should be identical. But we have not achieved this, yet. The demand side families are related to the products from the 2 plants, OTC products, licensed products, export, and more. The information distributed includes an analysis of actual sales vs. demand plan, showing percentage performance, for the last 3 months both individually and cumulatively. An error of 5 in 100, whether up or down, is shown as 95% accuracy; they were 5% wrong.
The demand planning meeting takes place around 2 weeks into the new month. It is attended by representatives from all stakeholders in the demand plan. Each demand manager presents to the meeting, family by family, the rationale for the proposed demand plan and the numbers are adjusted if necessary and agreed, and the Sales Director signs off on the new demand plan for each family. Risks and assumptions are clearly documented. We also turn the new sales plan into a revenue plan, and compare to the numbers in the budget and business plan. If necessary, we plan actions to get the
numbers back together, and these proposed actions get reviewed later on in the overall process. We look at the worst performing family, try to root cause the reasons for the poor performance, and put together an action plan. We monitor the achievement of our action plans. The agenda for the demand planning meeting forces us to look medium to long term. We spend very little time looking at the next 1 to 2 months; it does happen, but somebody quickly realises and gets us back on track. One of the issues we have not yet addressed is setting up ‘decision points’. For how many months forward should the demand plan be ‘frozen’ at the family level? Of course, when emergency situations, or opportunities, arise, plans can be reconsidered. For the next so many months forward, what level of change should we allow? Apparently, this is akin to the idea of ‘time fences’ in operations planning and the MPS.
We now have a consensus demand plan, into which all interested parties have had the opportunity to input. We used to worry about operations’ ability to meet our demand plan, and reduced the numbers accordingly. We now produce an unconstrained demand plan. This is what the market will buy, to the best of our belief. It is an open, honest, realistic, reasoned, reasonable plan. What we want to know is ‘can operations make it on-time, in full?’ Our plan shows both the new demand plan, by commercial family, and the differences from the old plan. We have to adjust the SKU numbers to match up with the family plans, but this is done by software, based upon percentages. So the end result is a new 18-month demand plan at both the SKU and commercial family level. Sales pass this to operations about 2 weeks into the new month.
3. Operations Planning
We now know what sales say they can sell. The question is; can we satisfy their demand on-time in full? We take the demand plan by SKU, and add it up by operations family. The operations families are split between the 2 plants, and are divided into white solids, colour solids and injectibles. We can therefore now see the differences between the old sales plan and the new sales plan, by operations family. This immediately gives us some idea of the likely impact upon operations and resources. We can see at this stage, just as we can in demand planning, the performance for the last 3 months. What was actual production by family vs. planned production by family, both for each of the last 3 months, and cumulatively over the last 3 months? Our target is 98% performance. We are not there yet, but we are getting there. We can identify the worst family, and we put in place some actions to improve performance. At this stage we can also see the actual consumption of key resources vs. planned load for each family, and in total, and likewise can identify the resource with the most variance, and try to root cause the reasons, and develop actions to improve in the future.
We match the new sales plan to the current operations plan, and look for mismatches where we cannot meet the demand. We have also established clear inventory holding policies for each family, defined in terms of months of inventory holding across all our DCs. So we can now see where the current operations plan and the new demand plan cause us to be above or below our inventory targets by family, over the 18-month horizon. We then start to make adjustments to the current operations plan, to make the numbers fall in line, generating our new operations plan, by family.
S&OP – How Did We Survive Without It? 4 of 7
The new operations plan is developed within known constraints. For example, we know the limit of the number of phials of injectibles we can produce per month at each plant, and we aim not to go above that limit. There are times, however, when we can see that the sales numbers, out in the future, are increasing significantly, and we know we have to plan to produce above that limit. We need to plan for this, so that, in the next step, we can see the resources whose supply needs to be increased, and likewise, those
resources which already have sufficient supply to meet the increased demand. We document our
assumptions and risks. When we have our first-cut new operations plan, we are ready to do our first pass at resource planning.
To avoid possible confusion, we prefer to use the term Resource Planning when we are talking about Sales & Operations Planning, and families. We use the term Rough Cut Capacity Planning when we are referring to the next level of detail, in considering the Master Production Schedule, the MPS, that we need to execute our operations plans, and to plan to make the individual SKU’s.
We have identified our key resources – the major and key resources we need to execute our family operations plans in full by the end of each month. We include here all our major departments in each of the 2 plants, so that we can see whether we have a capacity issue in our major departments over the next 18 months. We also include other critical, more detailed resources whose lead time is beyond the horizon of rough cut capacity planning, which is 5 months. We have 15 of these key resources. Some of them are aligned (to families), whilst some are non-aligned.
We have just 6 operations families. We are at the stage of considering breaking some of these families down to a sub-family level, and we will then conduct most of the S&OP process at the sub-family level, and summarise for senior management at the family level, with the details available when required. We have created a bill of resources for each family showing quantity of resource required, and the unit of measure; we have not needed to use offsets and durations, because our lead times are no more than 4 weeks. So for our key resources, we have a 6 by 14 matrix, stretching 18 months out into the future; we can grasp the implications quite readily, and make recommendations and decisions.
We developed some simple resource planning spreadsheets, which ‘multiply’ the bill of resources by the operations plan, for each period, for each family, and thus calculates a resource plan for each family. These are added together by resource to calculate total resource required for the next 18 months. We match this to our available capacity, both normal and maximum, and generate charts and graphs. From these we can quickly see which of our key resources are over-loaded, and can start to discuss and plan what solutions are needed. Where we do not have the resource available, and cannot obtain the resource to make the operations plan in the necessary timeframe, we work out what the maximum output can be, and where the demand cannot be satisfied. This is in preparation for the pre-S&OP meeting.
So we develop our best new operations plan to best meet the new demand plan. We identify the resource issues, and what the potential implications and solutions are. We identify where we cannot meet the demand in full by the end of each month. We are ready for the next stage in the process. It takes us the
3rd week of the month to come up with our plans. In the 4th week of the month, we need to integrate and
reconcile the numbers with the rest of the business, and we need senior management to review and approve the plans.
4. Integration and Reconciliation
At this stage of the process, product development, sales and marketing, and operations have been involved, but separately, and merely passing their plans to each other. Now we have to bring them together, to integrate their plans with each other, to integrate the plans with all of the rest of the business, and to reconcile the numbers with the higher level plans – the budget for the current year, and the ongoing business plan. This is all carried out in our RIM meeting – Reconciliation & Integration Meeting; this used to be called the pre-S&OP or Partnership Meeting, but we prefer the new term. Representatives from across the whole company, generally one level down from the Directors, get together to review the numbers, and to prepare for the Executive S&OP meeting. What do we do? In a smaller business, we suspect that this preliminary meeting would be overkill.
We start off by reviewing each family, looking at the issues and opportunities coming from product development, reviewing the implications of the new demand plan (looking at sales families), and
considering the issues raised by operations (reviewing by production family) in not being able to meet the demand plan, and what the resource issues are, and our plans to balance supply and demand to meet the inventory objectives.
S&OP – How Did We Survive Without It? 5 of 7
At this stage, finance comes into the equation. We can see the revenue generated by the new demand plan, and we match it to the budget and the business plan, looking for any gaps for which we need an action plan to propose to senior management. We can see the cost of the operations plan, at standard cost, and likewise compare to budget and business plan, looking for any excessive projected variances. Revenue minus cost gives us gross margin, and this too is matched to budget and business plan, looking for any gaps that we need to fill. This financial dimension to the S&OP process was something that we had completely missed before. We feel that the process is almost as much about financial planning as it is about sales and operations planning. Our financial people say they could not now live without our new S&OP process. It has even been suggested we call the process SOF planning, for Sales, Operations and Financial Planning! Get the ‘F’ word in there, to make the financial people realise the process is for them! We begin to review alternative strategies to achieve the budget and the business plan. We see whether the budget and business plan, are realistic sets of numbers that the company can achieve. We put together recommendations to the senior management team (SMT), if as a group we can agree on a proposal. If we disagree, we prepare alternative proposals, for the SMT to review and make decisions. We prepare the agenda for the SMT meeting, based upon a standard agenda, making additions or removals as necessary. During the meeting, all the major departments of the company can see the impacts upon them for the next 18 months. Do product development have to make changes to their NPD plans? Do sales & marketing have to develop new initiatives to raise the sales of a particular family? Do operations have to start planning for a new packing line to meet that level of sales? Do finance have to start considering the source for some new funds to develop that new product line? Do HR have to start planning the recruitment of 10 new people for the packing hall? Do IT have to start developing some new routines in the order processing software? All these have to be considered, and we develop proposals for
the SMT S&OP meeting. We have the RIM meeting at the beginning of the 4th week of the month, publish
the numbers, the agenda, and our proposals to the Executive, so that we can have the executive S&OP meeting before the close of the month, when, of course, we start the process all over again. We give the executive a couple of days to read and digest all the information, in preparation for their meeting.
5. Executive, SMT, Sales & Operations Planning meeting
This is scheduled for the last working Friday of the month. It is planned to be completed by lunchtime – usually it is, but we have spilled over into the afternoon, when we had a major decision to make. Because of the hard work done by our colleagues in developing all the numbers, we do not need to spend excessive time understanding what is in front of us. A couple of hours are usually enough. We only have 6 major families to review. We have maintained our long-term schedule for these meetings, and everyone attends. The agenda is relatively fixed. We look back at the action items from the last meeting, and update their status. We aim to keep the action items outstanding to a single A4 sheet, no more. Actions not completed in the month get carried forward, but turn to red on the minutes. We review the ’big picture’ of how the company is achieving against budget and strategic objectives. We take about 10 to 15 minutes to review each family – the new product status, if any, the current sales and marketing situation, the production and resource issues, the financial numbers, and any other implications. We review the recommendations made by the RIM meeting, and most of the time accept them. We consider the alternatives; we make decisions. If necessary we get the numbers reworked one more time to see the implications of another way of tackling an issue or opportunity.
We end up with a new NPD plan which product development commit to deliver. We have a new demand plan, which sales and marketing commit to sell. There is a new operations plan, which is achievable by the key resources, that production commit to make on-time, in full. Everyone commits to their part of the plans. There is horizontal agreement across the very top of the company to one set of numbers in one plan – we are a ‘one number company’. The attendees are now committed to communicate vertically down the company that one set of numbers. Everyone, yes everyone, finishes up working to one and the same plan.
The PA to the Chief Executive takes the minutes for the meeting. At the end, the minutes are reviewed and agreed. They are published within 1 day of the meeting. Our last step in the meeting is to consider what could have been improved in the process in the last month. We add one more action to try and ensure that we achieve continuous improvement in our SOF planning, month on month.
We are beginning to see benefits from our improved process, but we have plenty of progress still to make. The journey is a marathon, not a sprint.
S&OP – How Did We Survive Without It? 6 of 7
Benefits
We feel that the effort expended in the workshop, the lessons learnt, and the actions that we have now taken to introduce a real monthly S&OP process, are beginning to reap substantial benefits. It is too early to say that it has improved the bottom line, but all the indicators are pointing in the right direction. So what can we say we have achieved?
• We are now creating plans that reasonably balance demand and supply 18 months out into the future.
• Putting together the annual budget is no longer a major hassle; we start from the agreed sales and
operations plans for next year; we do it faster, so we have more time to concentrate on considering the options. The finance people consider this to be a major demonstrable benefit already delivered.
• We can use our longer-term plans for planning our material requirements out to the end of next year; we
can provide forecasts to suppliers, aiding our buyers in reaching even better agreements; these forecasts are more reliable than they were.
• We are now planning and controlling our inventory levels in a much more considered manner; we feel we
will achieve the numbers.
• The numbers that finance are working on are the same numbers as sales and operations, not a separate
set of numbers from a separate process. Finance are working to a set of current numbers, not to an old plan that was last reviewed up to 12 months ago.
• We plan our major and critical resource requirements 18 months into the future, so we had better budget for
capital acquisitions.
• We are working to realistic, achievable, doable plans.
• We are a ‘one number company’; where every employee is working to deliver the same set of numbers.
• We are measuring our performance at a higher, more relevant level, and our performance has started to
improve.
• The accuracy of our demand plans, our forecasts, is improving.
• We have a demand planning process that takes account of more of the relevant factors influencing our
future demand.
• The senior management can see the ‘big picture’ every month; we understand what is going on; we have
our finger on the pulse.
• Our medium to long term planning is significantly better than before; we would not yet claim our shorter
term planning processes are as good, but with better medium to long term plans we should be producing better short term plans; whilst our medium to long terms plans were poor, we were having to produce fixes in the short term, and sometimes failed.
• We have updated our policies and procedures in our quality manuals, so everything is up to date; new
starters can see the current processes, and can learn them from day 1.
• We are one happy team.
• We have made the right people accountable for the right processes and numbers; they are measured; they
know where they stand; they know when they need actions plans to get back on track.
S&OP – How Did We Survive Without It? 7 of 7
Conclusions
We thought we were doing quite well when we first implemented what we understood to be Sales & Operations Planning. Now we have (almost) done the job properly we are really feeling that we are in control, and we can feel and see that all the critical numbers are starting to move in the right direction. We now feel we have
processes in place that will allow us to hit our target for this year, and to become number 1 in our market sector. We cannot imagine now running the business without S&OP. How did we get through those early years of growth without it? We, for two, have no idea. But we will not rest on our laurels; we will strive for continuous improvement, both in S&OP and our other major processes, such as Demand Management, Master Scheduling and MRP.
Further Reading
We found the following material very useful and informative and would recommend anyone embarking on a similar journey to refer to them:
Title Author
Sales & Operations Planning, A Pocket Guide Robin Goodfellow
Sales & Operations Planning, The How-To Handbook Tom Wallace
Orchestrating Success Richard Ling & Walter Goddard
Enterprise Sales and Operations Planning George Palmatier with Colleen Crum
Demand Management Best Practices Colleen Crum with George Palmatier