For lean Six Sigma black belt skills, adding SQL to a practitioner’s tool set is a great complement to other lean Six Sigma tools that can be taught in lean Six Sigma training.
Most Lean Six Sigma projects end up working with data. We usually start with the data in a format that it is currently being used within the business.
This data is adequate to recognize the scope of the problem, but it is not usually sufficient to complete an improvement project. We need more data on the Xs (inputs) and other aspects of the business.
In my Black Belt and Master Black Belt (MBB) work, I have always have to fight to get all of the data I believe I need. This may take going out to collect data manually, but it is more often about pulling more data from the corporate data warehouse (corporate database).
To get the data requires a Database administrator (DBA) to pull the data or I do it myself. Doing it myself required me to learn SQL (structured query language). I believe all Belts should take time to learn SQL so that they can get their own data.
Additional Lean Six Sigma Black Belt Skills SQL: Reasons to do it yourself
- When someone else pulls the data by date range, you are never sure if it is all the requested data was provided.An example of getting the wrong data has repeatably occurred to me. I want all of the transactions from January. Seems like a reasonable ask, but in most cases the DBA provides me all of the transactions that both started and stopped in January. They leave out all of the earlier work that began in December or earlier and finished in January, along with leaving out the work that started in January and completed in February or later. The resulting data set is biased to not include longer task transactions.
- You request all transactions that involve client XYZ and you wonder if it is all of the data. It turns out that the request was case sensitive and when the client was entered as xyz the data was not included in the report.
- You want data on other process factors to see if they relate to the problem. The DBA gives you the multiple tables with the process data. These excel sheets are great, but how are you going to compare the results with Excel? Well it is impossible. SQL would make it much easier.
- You are provided a very large data set. It is near the maximum size that Excel or Minitab can open on your PC. Everything runs very slow. If you put the data in an SQL database, you can work with SQL to quickly extract data subsets and such.
What is SQL?
SQL is the basic language used to access data within relational databases. Relational Databases are what nearly every ERP software and corporate data warehouse used to store data. There are only 18 commands in SQL, it is quite a simple language. Of these 18, most of us only use 5. To use SQL, you do not need to be a programmer type. You just need to understand a bit of Boolean logic (and, or).
The basic commands are;
- Select: which is used to identify the data columns to output
- From: which is used to identify the data tables that the requested columns reside in.
- Where: which is used to describe how the tables are related and the specific conditions that must be met to provided an outputed value.
- Group by: which is used to provide summary data, such as sums, means, counts, and such.
- Order By: which is used to sort the output.
If you read my blogs you would figure out that I am an old school guy and like to do things the manual way to make sure I understand the process. If you do not want to use SQL you may recognize that Microsoft Access works with relational databases.
Microsoft Access has a very nice drag and drop graphical interface that generates the sql statements to obtain the data you wish to request. The down side of Access is that the SQL statements it generates do not follow the formatting as basic SQL. For me, I will play in Access and let it pull my data and generate my queries. Once I have what I want, I will take the Access SQL and simplify it so I can use it in other systems.
Are all SQL languages the same?
No they are not the same. It seems like every software vendor adds additional SQL functionality to the basic language. These additions do not transfer across SQL platforms. But the nice thing is that the basic SQL language does work on all platforms. So if you learn the basics, you can use it everywhere.
I learned to code in SQL on an Oracle database in 1990. What I learned there still works on the relational databases used today.
Where can I learn SQL?
Some of us can learn SQL by working with a current user and looking up questions on the internet. I think this works if you are already an analyst of big data sources.
Another method is to take a simple SQL class or read a book on SQL. I have found one class that seems to be simple and reasonable, but it is not free. It costs $25. It is by UDEMY SQL for Beginners.
Summary
Go learn SQL if you plan to be a Lean Six Sigma practitioner or a data analyst or if you are currently working in that roll. It is one skill that will pay a lot of dividends.