The nature of service work makes it difficult to find out what needs change and how to transform. The work product is often invisible making it difficult to track the flow. Service industry is has a long tradition of being individualistic. People are protective about losing their creativity to standardization. It is vital to engage the people instead of enforcing compliance. Data in service industry is neither organized nor readily available. More often than not decisions are judgment or Delphi based. People do not respond to inputs/instructions as do machines. The sheer unpredictability of human behavior is daunting.
Recognizing waste in service is critical to a successful LSS initiative:
• Over processing – PMI calls it “Gold Plating.” It may also include inefficient iterations of work product, unnecessary handoffs, creation of cumbersome documentation, etc.
• Transportation and motion – A lot of effort goes in information chasing. The data or the person having the data is difficult to reach out. The lack of resources may be result in information waiting at desktops for being processed.
• Inventory and waiting time – Non-value added work upstream increases downstream wait time. Skill and resource bottlenecks are also contributors
• Defect – Lost communication, miscommunication result in work product being incomplete or contrary to customer needs.
• Overproduction – Poor prioritization or work qualification leads to over commit and under deliver
Some useful ideas for running initiatives in service industry:
• Be creative in meetings – Use the meeting times wisely. Collect data offline. Use meetings to analyze information, make decisions, and have concrete takeaways.
• Look for obvious quick hit opportunities – Pick up low hanging fruit and use the results of those initiatives to promote initiatives in more critical areas.
• Use improvement events like Kaizen (traditional or improvised) to generate energy and immediate gains.
• Reach out beyond team boundaries.
• Set realistic expectation – A single projects does not produce best in class improvements. Processes are generally littered with variation and not managed with even a general understanding of velocity and flow. These are 2 distinct deficiencies and need to addressed separately
• Pay attention to team composition – Dr. Belbin’s model offers a good guideline, manage expectation
• Be conscious of your audience – Improvise tools and method to suit the organization, lead by example.
DMAICA typical LSS project begins with a one-page charter with sufficient data to compute benefits, resource requirements, and ROIC estimates. Basic elements of DEFINE include
• Consensus on the problem
• Understand the project link to corporate strategy and ROIC
• Agree on project boundaries/scope of value stream
• Know the key metrics or indicators of success (people must be able to make sense of it)
MEASURE bases on data and separates Six Sigma from a general process tinkering activity. Common data gathering challenges include too little data, no data, too much data, and irrelevant data. Useful steps in measure are
• Establish baseline – Measure useful data like things-in-process, average completion rate, cycle time, first pass yield, approvals/handoffs, downtime/learning curve, defects that affect the customer, and complexity
• Impartially observe the process – Watch people, track emails, phone calls, set goals for observation, ensure consistent data collection
The purpose of ANALYZE is to make sense of the data collected so far. The key part is to stick to the data and not be colored by individual experience and opinions. It may use scatter diagram to correlate variables. Time trap analysis can identify improvement areas.
IMPROVE applies standard tools implement brainstormed alternatives. The tools include those for setup time reduction (learning curve, upstream batching), 5S (De-cluttering), and queuing methods for decongestion (Staff pooling/cross training, task slotting/triaging, back up capacity, variation reduction). In a philosophical sense, nothing really changes except that the universe rearranges to be more efficient. Things like sponsor support and communication are of essence in this stage
A standard rule says we must spend time on a task only 10 times as much as it takes to set the task up. Some important questions for setup time reduction.
• What is delaying the start of work
• What factors lead to work interruption
• What inhibits moving at full throttle
• Is there any redundancy
• Is there a possibility for rationalization/ task offloading/ streamlining
• How to apply statistical control
The aim of CONTROL is to preserve the gains made in improve till yet another generation of improvement happens. Key steps are
• Documentation – to ensure people don’t slip back into old habits
• Convert results into currency – Project need financial validation and verification not be mere feel good factor. Not all gains are financially tangible. Verify the results at later points in time to ensure consistent returns
• Set up alerting mechanism to catch when the process spins out of control
Control is the cross over point when the process owner will take charge.
The book is an amazing read.
