“Bad Data” is often somewhat of a scapegoat when it comes to the failure of many supply chain optimization projects. However, when you explore this concept in more detail, you usually discover that it is a vague term that can hide a multitude of sins. In this episode of LokadTV, we explain why this is such an imprecise diagnostic and reveal some of the data challenges that can be encountered by both supply chain practitioners and software vendors alike.
In fact, “Bad Data” is an easy and convenient excuse because data can’t complain, it can’t defend itself and - unlike a colleague - it won’t take anything personally. But that doesn’t mean that the data isn’t without fault; data related problems are usually the number one cause for failure when it comes to supply chain optimization projects. Yet, there are numerous misconceptions.
For example, when we talk about “Bad Data” we usually think of corrupted data with incorrect numbers and typos, when really that’s not the main issue. For over a decade, thanks to the use of barcode and scanning technology, the vast majority of data is correct in this regard. So where do the real data problems lie?
Access to the data can be an issue, where old, out-of-date ERP systems are often to blame. Many of these systems are also not very “company friendly” when it comes to exporting and extracting data, some don’t even have a relational sequel database backing them up for example. Data integration can be another problem. Sometimes companies can even be placed in difficult positions due to conflicts of interests between the external IT company responsible for integrating the data and software vendors.
To wrap things up, we go into more detail about the key mistakes to avoid when handling and preparing supply chain related data - especially historical data - in order to be able to implement a supply chain optimization successfully. Supply chains are complex - this is unavoidable - with a richness of data that needs to be organised correctly if an optimization project is ever to work as intended.
00:31 Why do you think that ‘bad data’ is an easy excuse?
01:31 How can we be more precise?
03:06 What are some of the challenges we can encounter that result in data being not so good?
06:26 Changing an ERP system seems to be a headache for some of our clients. Which impact can this have upon the data?
09:45 What about forecasting? How easy is it to spot issues with historical data if you are using it to inform forecasts for the future?
12:19 Which issues does scalability introduce?
17:14 How widespread would you say those problems are?
24:08 What are the key lessons to learn from today’s episode? There is actually a wide range of different problems that come under the umbrella of bad data!