EDI produces a lot of data. Of course that's no shock to anyone looking at their VAN bill, but when you consider the volume of transactions and the number or companies involved, it would seem that there is a wealth of data in those transactions. The data now covers (depending on the trading partners) every aspect of the order process, from initial P.O. to final payment, with plenty of status updates along the way. So, what can be learned from all the data? It turns out that when looked at as an aggregate and put through the right analytical processes, there's plenty to be learned - and predicted.
EDI software/service providers/VANs that act as collecting points for EDI data are in a great position to help leverage this data because all the transactions they transfer between trading partners pass through their servers. At some point these transactions are stored on their servers, and some of the providers maintain those transactions for historical purposes. The newest trend that these providers are offering is to leverage those transactions by applying business intelligence techniques to them. What emerges from these advanced calculations takes on many forms, but in general they paint a picture of what has happened, and what is likely to happen in the future.
If you have not looked at my video interview with Chris Petersen of IMS here
, it will be worth a few minutes to see what Petersen and his company are doing with EDI data. IMS is involved in gathering EDI 852 (P.O.S.) transactions from retailers to help his manufacturer clients develop deep insights into how and where their products are performing. The company looks at trends by analyzing the register transactions from the manufacturer's retailer customers. As Petersen puts it, they are able to apply predictive analytics to adjust assortments and stocking levels in a near-time basis. He points to retailers that are advertising circular driven in particular, where an advertisement is delivered on the weekend, and data is analyzed as the sale is in progress. This allows the supplier to prepare for replenishment during the next week in order to meet demands in the specific outlets where sales are likely to deplete on hand stocks. The result is that sales are not missed because of out of stock conditions.
EDI service providers including EDIFice, Inovis
, SPS Commerce
, and Sterling Commerce
are providing various forms of data collection and analytical tools to their customers. I spoke with Erich Chaffee of SPS Commerce about that company's product, Trading Partner Intelligence (TPI). SPS Commerce collects transactions from its customers as part of its standard EDI connectivity services and offers TPI to its customers as a byproduct of this collection.
Chaffee explains that customers can subscribe to TPI, and because it's offered as part of the company's SaaS services, there is no installation or setup required. The results are presented in a series of web-based dashboards that are designed to be used by business users rather than by data analysts.
One of the main areas of concern addressed by this and other services of this type is the category referred to as 'order performance.' Companies are increasingly looking for the 'perfect order,' an order that matches the P.O. and the delivery exactly. By looking at the aggregated data, a manufacturer can see how they are doing with their trading partners in terms of delivering the perfect order. They are able to spot trends and identify patterns that may point out difficulties they are having with particular customers, locations, carriers, or products.
Being able to use the data created by both the supplier and customer to understand the sales cycle can be a great advantage to both companies who are looking to make the most of their resources in an increasingly competitive environment. As Petersen explains, the addition of CPFR (Cooperative Planning, Forecasting, and Replenishment) can go a long way to maximizing margins and minimizing mistakes.
The key word here is 'Cooperative' but the tools that make this process possible, if not easy to implement are already available because of the collection of data that makes up the EDI transactions that till recently, have moved through the supply chain with little analysis.