00:04 Introduction
00:22 Why is it that the data may not be appropriate for working with Lokad?
01:39 What is the minimum time frame of history that you should have?
04:24 We often talk about the four main datasets that we require. Let’s start off with the catalogue, what’s interesting here?
07:06 With the sales history, what other information do we need other than what has been ordered, where, when and by who?
11:16 Other than knowing how much we are paying historically for our goods, what other key information can we generate from purchase orders?
16:54 Can we not just use our sales and purchase history to generate the stock level?
20:54 We have spoken a lot about the main core data that is required. What about the other data that can be included?
23:52 What can we learn from the companies that Lokad deal with on a daily basis? What common mistakes do they make?
26:25 If there was one thing to take away from today’s discussion, what would that be?
Description
Predictive supply chain optimization relies on heavily prepared data. The purpose of this data is twofold: first, the historical supply chain data is used to build the forecasting models, second, the data describing the supply chain’s current state is used to drive the optimization of the decisions.
Not all data is appropriate for a Quantitative Supply Chain initiative to be effective. In this episode of LokadTV, we understand what data a customer needs in order to work with Lokad and how best to record it. We investigate the depth of data that our clients actually require and also the common mistakes and pitfalls that can be encountered along the way in collecting this data.
The four main datasets required by Lokad are: catalogue, sales history, purchase orders and current stock levels. We go into more depth to understand just why the catalog is of primary importance for us, and how our technology can leverage the correlations between products for better forecasts. Concerning the sales history, we explore which are the mistakes to avoid in order not to have an incomplete sales history and the importance of not losing track of the products that the company doesn’t sell anymore.
We then talk in more detail about the purchase orders, and the information related to this that you should record, such as quality issues with the order for example. In addition, we focus on the stock levels and the necessity to record not only the current stock levels but also the historical ones (with a particular emphasis on stockouts).
We expand on the minimum time frame of depth of history needed for a supply chain optimization initiative to be effective. To wrap things up, we try and learn from the mistakes that companies working with us make on a daily basis: for example, choosing an ERP that makes the data you need super difficult to extract…
