Improve

There have been a number of different approaches to process improvement over the years. One of the earlier ones was Total Quality Management (TQM), which was later followed by Six Sigma and Lean. They all aim to provide continuous improvement, however their focus is slightly different. Despite this, the fundamental approach is common, with the differences being in the methods and tools employed. To avoid labels, my summary is agnostic and describes the fundamentals. Although changes to existing processes should always be done with caution, that does not mean that all improvement will take a long time.

 
Move fast 1.gif
 

Whatever the type or complexity of the process, its improvement should always include answering four questions. Fortunately, there is no need to be a Zen master, as the questions are deceptively simple but the answers will require some thought.

  • What are you trying to accomplish?

    Even though you may have a good idea of what needs to be improved, it's always worth defining the problem carefully, otherwise the outcome may not be what you wanted. Use "Cause > Problem > Consequence" to position the true problem. Think about how it affects your own organisation, any external suppliers and most important of all, your customers. Define the scope, so that you don't accidently include the whole organisation, and include specific measurable objectives, so that you'll know when you've reached them. This leads on to question number two.

  • How will you know if a change has been an improvement?

    This is where you'll need some measurements, both for your current state (a baseline) and intended future state (your objectives). When deciding what data to capture, don't forget what's important to your customers. Also consider that you may make a change to one process that will affect other processes in your organisation, so you may need to include some wider ranging measurements to capture any of these unexpected changes. The Linkage of Process map is useful for this purpose.

  • What changes are you going to make that will result in an improvement?

    Whether collecting data, analysing the results or making changes, use a learning technique called the Plan > Do > Check > Act cycle (PDCA), which simply means that you plan to do something, do it, check the results and decide what to do next based on what you found. Your problem may require just a single cycle or many cycles.

Before changing anything, the first step is to analyse the current state, if not already known, with a view to finding one or more root causes. There are two different techniques that are commonly used, one of which is good for quality problems, where you might have a process output varying beyond acceptable limits and you need to find the cause, and the other is good for maximising productivity and minimising waste in its various forms, usually in a chain of processes across an organisation. It can be a case of using either or both, at different times. Click on the links below to learn more.

Both techniques originated in manufacturing but have since been applied to many types of organisations and environments, so they are equally at home in a factory, a hospital, a call centre or an office.

  • How will you ensure that any improvement is maintained over time?

    This is all about the long term management of the process(es) which have been improved and unless this is done you run the risk of being back where you started within six months. Following the updating and standardising of any changed processes, it is best to assign people to the ongoing upkeep of the improved process(es) as part of normal business. This will usually involve recording and monitoring key metrics and ensuring corrective action is taken when required.

Tracking Progress

The usual project plan format is not entirely suitable for an improvement project, as although the desired goals are stated at the start, the exact method of reaching them is unknown and will evolve as the exploration progresses. There are at least two basic formats in common use.

  • The first is often referred to as an A3 report since it illustrates a complete project on a single side of A3 paper, using the PDCA model. It was pioneered by Toyota as a concise method for engaging all teams throughout a project and for gaining their consensus. It is very suited to fairly short projects where the scope is well defined.

  • The second consists of an overall container for the above questions and subsequent answers, together with specific sections for the DMAIC (Define, Measure, Analyse, Improve, Control) phases defined by the Six Sigma method. The PDCA cycles used in each of the phases are either included or referenced separately. This format is quite complex, compared to the A3, but is suitable for larger projects where the DMAIC model is used.

A Cautionary Tale

Some process improvements fail to meet their objectives and these are some of the common causes of failure. Knowing of them in advance may help you to avoid them.

  • Processes selected are not simple and the effort and commitment to improve them is underestimated.

  • Objectives and scope are not clearly defined. There is a temptation to improve the whole organisation.

  • Measures not defined or appropriately linked to stated goals.

  • Insufficient management buy-in, so in-house teams do not have the time required to actively participate, nor do they feel ownership of any changes made.

  • No effective root cause analysis, so proposed solutions may be solving the wrong defined cause.

  • Poor change management for the agreed changes, especially around training and documentation.

  • No follow up to ensure improvements are maintained over time.