Supply Chain Leaders: the empirical approach

Date : 01 May 2019

With part two of our Supply Chain Leaders series published last week, I thought it would be a great time to explore further how you can embrace the empirical approach in your journey to supply chain excellence.

Our Supply Chain Leaders work has improved our understanding of how supply chain is changing, and the attributes needed to succeed. The first attribute explored in depth is empirical.

An empirical approach emerged from the survey results, with leaders’ recognising the role data plays in their supply chain, but also the growing importance of experience – two foundations of empirical working.

What it means to be empirical

To be empirical, we can’t make decisions based on theories or intuition alone. What comes out of these tests is empirical data – the objective facts you need to make informed decisions. You may also decide to make use of existing data to verify a theory – let’s face it, it’s sometimes easier and there’s plenty of it available! These are both legitimate empirical approaches.

Being empirical is getting easier

The survey results tell us that conditions for empirical working are improving – there’s more data available to test theory and it’s easier to set up experiments to create new data. But data is only half of the story. Data must sit alongside experience, which should be used to first identify what to test, and where the likely value is, and second, to interpret and bring meaning to results and help determine how best to act on them.

Doing so should be our goal because the consequences for getting things wrong can be severe – wasted time, eroded margins and ceded advantage are all avoidable potential consequences of not testing our ideas. The report outlines five things you can do to embody the empirical approach:

Source: IGD Supply Chain Analysis, Supply Chain Leaders Pt II

Make use of what’s available

I’ve mentioned that the conditions for empirical working are improving. A key driving force is the shift towards digitalisation and automation. 83% of respondents stated that automating processes was a key part of their company’s strategy or it is their number one priority. But, we do need to be careful. Automation and digitalisation will mean machines, in addition to humans, will be using data. Algorithms must use data that’s “tested” and machine learning, in which algorithms update automatically based on results of the previous decisions made, can support this requirement.

This continuous learning loop is needed to ensure we continue to be empirical in a world where humans make fewer of the decisions and are less able to rely on our personal experiences.

In the meantime, platforms like Metro Group’s Supplier Collaboration Tool (SCOT) will remain  crucial components. It provides manufacturers with near real-time data on Metro operations through a 24/7 self-service platform.

Data from platforms like SCOT can be used to check the veracity theories and gather evidence, and can be a great starting point, helping you move from theory to a robust empirical assessment.

Using existing data is a great way to identify opportunities. This is certainly a valid empirical approach but, of course, there’s no guarantee the data you need will be available. In which case, you may need to generate some.

Simulate and experiment

You can do this quite easily and it really doesn’t need to be complicated. We’re seeing “split” or “A/B” testing increasingly used to measure theories. For example, “adding a delivery day to customers’ order schedules will smooth the order profile, making it less erratic”. An A/B is perfect for testing theories like this without committing to action that could drive negative outcomes.

To measure such a theory, an A/B test is ideal and empirically sound. In this example, it would mean allowing a control and a variation, which simulates the conditions outlined in your theory, to take place simultaneously. The results can then be assessed, and the empirical data used to decide how to proceed.

Don’t be daunted by the thought of statistical analysis, our change management tool can help you!

Running an A/B test

 

Source: Supply Chain Analysis, Change Management Guide

The tool walks you through how to run an empirical A/B test and provides a template that you can add data to.

Conclusions

Being empirical will probably change how and where you spend your time. It calls for more upfront investment to reduce time spent late, but in doing so, ensures you don’t waste time and do focus your efforts on the things that can make the most difference.

If you haven’t already, take a look at our first and second parts of the Supply Chain Leaders series.  Look out for the next report, which explores the next of the Five E’s – entrepreneurial.

Alex Edge

Alex Edge

Supply Chain Insight Manager

Download our report to understand how supply chain excellence will be a source of growth and value for the future.

Get the latest industry news and insights straight to your inbox with our range of newsletters.