The objective of supply chain integration is to make better decisions with less risk by receiving timely information about supply and demand factors occurring up and down-stream. Ideally, this would involve IT system-level integration of companies with all their suppliers and customers for real time access to data in both directions.

Companies don't need to pull supply chain information if their suppliers and customers are willing to push itUnfortunately, multiple significant barriers often keep companies from achieving that ideal:

  • Security concerns of companies who would have to grant access.
  • Competing demands for the resources needed within companies, their suppliers, and customers to define and accomplish IT-level integrations.
  • Difficulties establishing communications among systems that vary considerably in operating systems, communication protocols, and architecture.
  • The sheer number of parties that could be involved in any given month or year.

However, even if companies can’t achieve the ideal of being able to pull relevant data from all their suppliers and customers, they can achieve many of the benefits by having suppliers and customers cooperate in an “info-push” system. In an info-push system, suppliers and customers periodically send (i.e., “push”) reports with the desired information, and the receiving company recreates or reverse engineers the data structure upon which those reports are based, and consolidates the data in its own internal database.

There are two key concepts involved here:

  1. Acceptance and cooperation. Suppliers and customers will often be much more willing to create and send reports than they would to permit third parties to login and pull that information from their systems. The reason: they don’t have to worry about things like VPN tunnels, security protocols, SQL injections, and the like when just sending reports.
  2. Consistent input. Information in the reports sent by suppliers and customers is created by database systems, and the reports are quite consistent in terms of layout and content. Accurately extracting information from such reports is a case of high value with only modest effort.

Info-Push Example

Here’s a simplified example of how this could work: A company has 150 distributors of its products. Some distributors are large and highly automated, but many of the smaller ones aren’t. Most of the automated ones are unwilling to provide the access and the resources to achieve full supply chain integration. However, just about all of the automated ones are willing to set up weekly inventory reports for the company’s products and automatically email them to the company.

The receiving company can use technology to detect the name of the sending party, and map the data on each report to specific fields in an internal database, even if each of the automated reports uses different layouts and different column and row headings. The identification, extraction, and loading of information can be completely automated.

The advantages of such a reverse engineered and consolidated database are significant:

  • Custom reports. With consolidated data, companies can present data from different suppliers or customers in a standard format and not have to compare what may be apples to oranges.
  • Trend lines. Accumulating information over time lets a company view trends that may not be obvious from isolated reports from individual suppliers or customers. Having data from many suppliers can provide insight into different market segments and geographical areas even if all suppliers don’t participate.
  • More Efficient distribution of data and use of time. Rather than requiring multiple people in a company to monitor all incoming reports or notices, having a consolidated database enables the use of automatic notifications when triggering events occur. A database would also serve as a single resource to answer questions from information in the reports.
  • Model or substitute for actual IT-level integration. Existing reports and notices received from suppliers and customers are good starting points for defining what information would be desirable for full integration. Having a reverse-engineered and consolidated database can also inform the decision of whether full IT-level integration is actually needed – there may not be much incremental gain from full integration. The consolidated database can also be used to supplement a fully integrated database to manage data from suppliers or customers who don’t participate in the full integration.

The internal database would not be real time, but real-time data may not be essential. For example, for decisions made weekly, most data provided hourly or daily are ignored.

BeyondRecognition’s visual classification and attribute extraction technology offers one way to quickly classify incoming information, extract key data from it, and feed a database that consolidates all the incoming data from suppliers and customers.


For your free personal copy of my book, Guide to Managing Unstructured Content, go to: http://beyondrecognition.net/download-john-martins-guide-to-managing-unstructured-content/

Comments are closed.