In our previous discussions I introduced concepts around how to start organizational change from a
combined bottom up/top down approach. We have discussed how to introduce an initiative which focuses on improving the organization in terms of value delivery. We have also discussed how to mind map the organization to find troubles, constraints and a true picture of your company. Once your problems have been identified it can be easy to try to simply "fix" the problem. What I mean is that human tendency is to correct what we think is the biggest perceived symptom. Approaches of this nature commonly lacks greater vision and likely only moves the problems or defers to another bottleneck. When this shortsighted plan just defers to the next constraint the whole effort is often construed as a failure. What must be considered when facing organizational change is the value stream as a whole. Understand as many constraints or non-value added activities as possible when mapping the stream. In all my consulting experience I have never seen an organization who begins a value stream map exceeding 6% of value added activity. Setting expectations with leaders from the start to have patience is a key to organizational improvement.
I was working with a development team who believed they needed to hire more developers because it was taking six months to get code released (forget agile for the moment)to production. This was their problem statement (and their perceived solution). After creating a value stream map of the development cycle we learned that the cycle time was almost 900 working hours. Four hours of that time was spent developing. Roughly 80% of that time was wait time. Had the team hired more developers they could have attempted to double their productivity and take two hours out of their total cycle time. Wow what an expensive solution. Or they could have eliminated several management meetings which were simply rubber stamp approvals to move forward. Reducing the meetings took 300 hours off of the total cycle time, and it cost nothing to implement. To get that efficiency gain by hiring developers is simply impossible yet this is most commonly the first solution to reduce development value stream times.
The concept of value stream mapping has been written very extensively and while there are many good posts most simply introduce the concept but fail at plugging the concept in to something actionable. Or they prescribe perfect ways to change your value stream without providing a lesson on why focusing on the value stream is relevant. I will refer to many smarter practitioners through this blog but it's important to define what value stream mapping can accomplish in terms of starting organizational change. The LSS Academy has put it best at defining a value stream as "A value stream can be defined as all the steps – both value added and non value added – required to take a product or service from its raw materials state into the waiting arms of a happy customer." Again the concept of value in this scenario is simply what your customer is willing to pay for. We know that there are Value Added (VA) activities which the customer finds worth to pay for. Non-Value Added (NVA) activities are usually where great debate begins. This work is often very overwhelming and it's not a bad idea to look for an outside consultant to help you. This is really the only time where I encourage going outside to make this success.
Again it's important to remember why we are doing all these activities. Shigeo Shingo said, “The four goals of improvement must be to make things easier, better, faster, and cheaper.” Note the sequence of this list; Lean focuses on adding value and eliminating waste through simplification, quality improvement, and lead time reduction; cost reduction is a natural outcome. According to Gartner, “One of the biggest mistakes companies make with Lean is focusing too heavily on driving cost out of the business. If you approach Lean with that attitude, you’re bound to fail. Building a huge data center may give you great economies of scale, for example, but it may also reduce agility and flexibility.”(Bell, Steven C. (2012-01-04). Lean IT: Enabling and Sustaining Your Lean Transformation)
Consider your own organization, you are likely structured in the form of departments which scale and report vertically. Who is responsible for the product from inception to cash? A value stream is the horizontal flow of your product from that inception to cash. This will feel odd to your organization because they rarely care about their products in that manner.
The best analogy is to compare the value stream to a story. A product begins its journey in marketing and end in accounts payable. The key to creating an excellent current state VSM is to document what you actually see with your own eyes. We are not interested in how the process is supposed to work, or was designed to work. Instead, we are interested in how the process is performing today. Will the process change a bit tomorrow? Sure. But that’s OK. In fact it's encouraged. There a many different ways to create a value stream map and my personal opinion is that you use one that works for you. As long as the result of the exercise is to identify your waste or NVA activities in your stream. Below is a good example I found from ASQ on a simple value stream map.
Now lets get started. The teams created for the prior steps in this journey should also be used to value stream map. They will understand the mapping and problems in the organization best and it's encouraged you stick with those team members. They are likely the only people who have ever looked at the organization horizontally versus vertically. For example I once worked with a large insurance organization who were structured in to operating companies. No big surprise there but each OP Co. was designed to support certain business segments (home insurance, life insurance, investments). Within each segment we also saw turf wars where they hesitated in helping other departments. Activities as easy as sharing customer information or transferring phone calls from the customer were almost impossible to accomplish. Where's the value in this line of thinking? Well it's easy to keep that mind set when you only look within you own department. Encourage your people to think across the value stream and you will see very different outcomes.
1. Draw the current state value stream map with all of the steps, delays, and information flows required to deliver the target service. IF you're interested in some templates please contact me.
2. Assess the current state value stream map in terms of flow and waste. Collect all missing data (e.g., value added, cycle time, etc.).
3. Calculate the total lead time and value added time.4. Identify and list all areas of waste (e.g., inventory, transportation, etc.).
5. Develop a list of opportunities for improvement based on these observations.
Below is a great example of mind mapping a development process (From the blog Wide Awake Developers, http://www.michaelnygard.com/). This can show a good example of how Agile and Lean can fit together in the IT world. Think of an agile activity as something which could exists more at an execution level in each activity while the Lean Value Stream focuses on the end to end process.
I’m going to use the prescribed method from the LSS Academy below. Now this has a manufacturing focus but can easily be replaced with services or code.
Walk the process front to back. Quickly walk the process with your team in order to understand the general flow. It’s important to also define the start and stop point of the process. Don’t attempt to take on too much. Remember, we eat an elephant one bite at a time.
Draw in the customer box / details. In the top right hand side of the paper we draw the little saw topped box representing our customer.
We also note their monthly and/or daily demand along with the takt time as calculated in step
Go to the end! Next, we start at the END of the process and begin drawing the map back to front. And don’t forget about that eraser. You will need it. Nominiate a scribe and have them draw the map for the team.
Another trick is to ask each person on the team to map it out so you can compare and consolidate when you get back to the room. Yet another trick is to have the team divide and conquer as you send some off to map the beginning section, some to the middle, and some to the end. There are many ways to do this. Experiment and do what works best for your situation.
Focus on the flow first. Focus on the flow side of things first (bottom portion of the map). This includes the process boxes and data boxes. Regarding the data boxes, if you don’t have all the data perfectly collected on the day of the mapping exercise just do the best you can. You can always assign homework to go back and validate the figures later. In fact, even if you think you have solid data, the six sigma side of me urges you to validate your measurement systems to make sure we can trust the data. If you want to get really tricky state both a measure of central tendency and dispersion. You won’t see this advice in most lean VSM books… I guarantee it!
After studying the KB&R manufacturing process for an afternoon we learned that each process step is staffed with 1 operator. We also collected cycle time information at each step. Additional “homework” will be to collect information such as defect rates and changeover times.
Add the Inventory/Wait Times. Once you have all the process and data boxes in, it’s time to add in inventory and/or waiting times. For inventory, we simply count the number of components in between the processes and note them under the triangle.
We also want to convert these components into days’ supply. To do this, we divide the number of components by the average daily demand (which we used to calculate takt time).
So, if your average daily demand is 10 components and you count 20 components of inventory in between process step A and process step B you have 2 days’ supply (20/10) in between the two processes. We will note this number on our timeline (to be added in a future step).
Lastly, don’t attempt to map every part number! Choose one or two key components to start with. You can always add more to the map later. In our example, we chose to simply count two components of bread as one subassembly since they move together down the production line.
Also, we are not accounting for the peanut butter and jelly “raw material” at this point since KB&R’s expert supply chain team negotiated a killer consignment stock deal with Sam’s Club so this inventory is quite low on the line. During the study, we learned that, as one example, there were 486 sub-assemblies (972 components of bread) in between the jelly application and packaging stations. This equates to 0.69 days’ supply (486 units / 700 daily demand).
Lastly, during the walk through of the process we noticed that each process step seemed to be working in isolation. In other words, the lady working at the peanut butter application seemed to produce as many units as she could and then pushed them along to the jelly application process.
This “push” process is found in just about every mass production process known to mankind. When we see this pushing action we note it on a VSM with a dashed line through the yellow inventory symbol.
Draw in the information flow. This step is what really separates a VSM from traditional process maps in my opinion. You see, in addition to learning about how material flows we also want to understand how information flows.
For example, we want to know it is moves about electronically? If so, we use a lightning bolt looking arrowed line. Is it communicated manually? If so, we use a straight arrowed line.
During this step we also draw in our production control box. For many, this box will include the letters “MRP” in it. In most mass production systems we typically see several manual information (straight) lines coming out of the MRP box aimed straight at each process step box.
In our example, we learned that production schedules each process step in isolation. In other words, each work station gets its unique production schedule. We draw this using straight “manual” information lines.
We also add in the information flow from our customers as well as to our suppliers. In our example, we learned that PB&J’s customer sends 30 days electronic forecasts as well as electronic daily orders. Conversely, PB&J sends its bread supplier an electronic weekly forecast.
Add in the timeline. We can now add the timeline to the bottom of the value stream map. This saw tooth looking line helps us separate the value added cycle time (taken from data boxes) from the non-value added time (days’ or hours’ supply info).
The last step in the process is to sum up all the “value-add” cycle times and note them at the end of the timeline. Likewise, we also sum up the “inventory” times and note that on the timeline.
Points to keep in mind.
Prepare for the High Cost of Participation. While value steam mapping can be an effective process improvement tool, the cost to use it can be high. Like racing a top fuel dragster, value steam mapping requires the presence of highly skilled support team members who practice frequently. Because of the need for considerable detail, team members will spend hours, even days, to develop a comprehensive value steam map.
Unlike drag racing, where direct costs can be measured, value stream mapping involves the risk of opportunity costs that are difficult to quantify. Simpler, more productive tools that are easier to learn might be available to make describing the current process faster and more effective. Many organizations forget to consider these opportunity costs for using value stream mapping.
This take times. Remember if you looking for a quick fix you might find it here but will quickly be disappointed when the entire effort fails. This is not only a change to your process but more importantly it’s a change to your culture. People take time to change. If your leaders aren’t supporting this from the top down and in the long run then you’re just going to spin your wheels.
Other Problems will appear. If you’re familiar with the Theory of Constraints then you know that anytime one bottleneck is identified and eliminated another one will appear. That bottleneck as always been there but has not been your primary constraint. Don’t be discouraged if you find a mind field of bottlenecks. It’s proof that what you’re doing is working.