Lean Techniques for Data Quality Management: Little's Law
by Dylan Jones
In this article we look at another Lean tool that works perfectly with data quality improvement initiatives: Little's Law.
This simple tool enables us to calculate the lead time for a business service. This in turn allows us to quickly identify areas where effective data quality management will help us deliver faster services, reduce costs and most importantly - delight customers.
If you're looking to accelerate the time-to-benefit of your data quality initiative this may just be the vital ingredient your're looking for.
What is Lead Time and why is it so important?
Lead time is the length of time it typically takes a business to complete a specific activity that is of value to the customer.
Imagine you order a new CD off the internet, the lead time is the time taken from placing the order until it arriving at your door.
It's probably one of the most important metrics your organisation has for benchmarking its service against the needs of its customers and its competitors because it provides two things:
- A promise or commitment that is measured by the customer
- A metric that consumers use to compare against your competitors
Let me explain with a recent personal example.
We are currently moving house and need to purchase various items of furniture and electrical equipment very quickly.
Browsing on the internet we can see all the different goods to choose from but for us the most important thing is that we get the item delivered quickly and on a specific day, anything else causes us real problems.
The variation in lead times is quite marked between the different vendors. Some deliver in 2 days, some in 2 weeks, some in 3-10 days.
So for us, the accuracy and performance of the lead time are the most important buying factors and it's the metric we use most to compare vendors.
Price is important because we want to get a good deal but if the freezer turns up 5 days late then we've effectively lost any savings as our food will already be destroyed.
In this modern world where we want everything instantly, reducing your lead time is vital to success.
What Lead Times are Critical to Customer Service in Your Organisation?
In the last Lean tutorial we looked at the importance of Time Value Maps, a great tool for determining where in a process to focus your data quality efforts for maximum gain.
The most obvious lead times are those of customer purchases so can you deliver as fast and as reliably as your competitors? Does your business regularly benchmark its own lead times and look for areas of improvement?
It is surprising how many companies have no idea of their own performance in the market so take a hard look at your organisation and identify the customer service processes that may require lead time improvement.
What is the Problem with Lead Times?
There is one problem however, particularly in large organisations, which is where do we start looking?
There could be scores of different service processes so we need a simple tool to help us identify those processes that are suffering from larger lead times than our competitors.
But there is another even bigger problem.
We can't spare the resources to trace each service chain to calculate its duration. Sometimes these can take many months to complete.
So how do we find these lead times quickly?
The Solution to Data Quality Focus - Little's Law
Little's Law, named after the mathematician who proved it, provides the average lead time by examining how items of work are still in progress and how many are typically completed in a particular period.
The equation is as follows:
Lead Time = Amount of Work-In-Progress / Average Completion Rate
For example, imagine that a firm completes 10 widgets a week and the current number of widgets being worked on is 100. The average lead time for a customer to receive their widget is therefore 10 weeks.
This equation, although simple and rather obvious, clearly demonstrates why poor data quality can have such a dramatic impact on lead time.
If we have poorly designed information chains, frequent order handling data entry errors or any of the myriad of data quality issues that plague our customer handling and provision processes, we are simply keeping more Work In Progress (WIP). By reducing WIP we actually make the process go faster!
There is a significant chance that by combining Little's Law with our Time Value Maps you will find exactly:
- Which service chains have poorly performing lead times against customer needs
- Which service chain activities have the greatest areas of waste
This process is clearly not rocket science. It doesn't need a Six Sigma black-belt or management consultancy to perform this kind of activity and any size of organisation will benefit from lead time acceleration because it delights customers and reduces costs.
You can typically find a huge amount of "low-hanging fruit" to pick that is adding WIP through the past introduction of batch processes, mistakes in data entry, insufficient training and poorly managed information chains in general. Find these data quality quick-wins and eliminate them from the service chain.
Still not convinced?
I've personally seen this exercise strip 4 months off a 5 month lead time simply by:
- Creating a data quality firewall on inbound data to eliminate defects entering the process
- Training all data entry staff in basic data quality (they loved it and all went on to become data quality/data analysts, morale rocketed)
- Elimination of all batch processes
- Removal of bottlenecks with key personnel
Data quality is not about cleansing data or profiling columns. It's about delighting your customers through faster, cheaper services that offer the same quality of service time after time. By eliminating waste in the most important services you will also create less variation in the quality of your service and added value in the eyes of the consumer.
Consumers Hate Variation in Service - You Can Use This
Have you ever waited for an appliance to be delivered that never arrived? Did you choose that vendor again? Did you complain to your friends at work after wasting your annual leave to wait for a delivery person who failed to show?
So don't delay, go find those services that have lead times that matter to your customers and find those simple data quality improvements that really matter.
By making instant improvements in the most focused of areas you can start to gain immediate results for your data quality efforts.
Image credits: Creative Commons SteveD