Baselining a Project
Nearly almost all Lean Six Sigma projects that experience a successful outcome are based on an understandable, quantifiable, agreed-upon, data-collectable baseline. Lewis Carroll gave us the heads-up in Alice in Wonderland and George Harrison, the ex-Beatle, later sang this verse in his song Any Road: “if you don’t know where you are going, any road will take you there”. Of course, Lean Six Sigma professionals do want to know where they’re going, especially when it comes to deploying successful projects.
Investigating, validating, and setting a good baseline will help you guide your project with a constant “eyes on the prize” type of approach. But how do we go about it?
1. Critically seek out the “metric of truth”: what exactly are we trying to change here? How do we know if we will change it or not? Where does the data come from, and how do we acquire it?
2. Fearlessly dig deep into the available data and diligently study the link between the data and the metric you are trying to establish – does it make sense? For some processes, this is given:
- a piece of machinery that measure its process in inches
- a sound survey that measures the favorability index of a group of customers in percentages
- a dollar amount (savings) that can be easily linked to waste elimination in a food manufacturing plant
However, for some other processes, this exercise might not be that easy and mathematical in nature, but the point remains the same – how do we measure the impact of the project?
3. Draw insights from your baseline data before you throw an average number on the project charter. Do you see clear special causes of variation? Do you see unusual outliers? What about seasonality? Any abrupt change in patterns? Any obvious shifts or trends?
4. With your Project Team, define the best period of time in which the baseline data best represent the current process – this varies widely amongst industries, but a good rule of thumb is that your baseline will clearly showcase how the improvements to be made will actually change your metric of choice. You don’t want data that represent a decade-old process, and you don’t want data that was collected in yesterday’s morning shift either.
5. Review the assumptions and make sure that all stakeholders are aware of them. For example, consider an expected level of inflation if savings are related to raw materials costs. Assume that labor shortages are not likely to occur in the next six months or that the supply of a certain chemical will not impact production – things like that. When setting up your baseline, most likely your CT (Critical To) metric will be based on a few variables so make sure that these variables cannot all just change at the same time – otherwise your outcome will always depend on things that you cannot control rather than on the operational metric under improvement efforts.
Start your project with profound knowledge of the process and its metrics. Ideally, do the analysis to set the baseline. But, avoid “analysis paralysis”. Plan well, do your homework on baselining the project, and move along – at the end of the day, a road with good signs of a successful arrival is better than “any road” to take you there!
I love Continuous Improvement and Data Analytics. The world would be a better place if our kids were taught more process excellence and statistical analysis at school.