A well planned e-commerce integration and whether to choose a middleware package over a single silo EDI solution is closely related. Before I begin to examine this relationship it is important to define a couple of key terms.
EDI: structured messaging as defined by standards bodies like ASC X12, UN EDIFACT, TRADACOMS etc.
Electronic messaging: any form of structured electronic message that can be consistently transformed. This would include various message structures such as XML, JSON, flat files, CSV and of course standards based EDI.
Middleware: an integrated suite of messaging and transformation tools designed to handle multiple communications strategies and multiple message formats. One of the key middleware responsibilities is to impose order and control over disparate messaging origins and destinations
Integrated environment: Is one where input flows from various sources in a variety of message formats. Data flows in and flows out to all of the necessary internal systems as well as external systems, without manual human interaction.
Now that the definitions are out there and we all speak the same language let’s examine the relationship between a well-planned system and choosing a middleware package
A truly integrated environment, where incoming messaging inputs kick off a series of automated processes and supply chain decisions, will reap the greatest benefits. A truly integrated environment will not come without a lot of work and fine-tuning as data flows.
Much like the final fully integrated environment,the data flow planning requires a great deal of time working through various cause and effect scenarios. Flow relationships between applications and repositories need to be established. This requires a considerable effort to make the system transparent, robust and auditable. Exception handling often takes as much time as basic flow. The larger the organization; the more complex the integration; the more time and the more money it will take. So the buy in, in establishing a fully integrated environment is to establish a clean and concise data usage strategy.
Today, systems are no longer confined to an internal local area networks. Messages no longer come from a few select supply chain partners, instead we have web stores, mobile apps, and cloud-based services to contend with. This list is not shrinking, in fact it is only growing and the corresponding data touch points and integrations are getting more and more complex.
Fortunately, standard integration methods have proliferated and matured at the same time. Today, it is more commonplace to ask if the application or cloud-based service has an API rather than can I FTP files or communicate with a VAN.
In this context of ever growing complexity (and opportunity) it is important to keep “future proofing” in mind as much as possible in both planning as well as the integration platform you choose.
This leads us to the use of middleware. Middleware as I mentioned earlier allows one to leverage an integrated suite of messaging and transformation tools. This allows for flexibility and speed in responding to integration requests and partner onboarding. It helps create an auditable and searchable document trail. If you choose feature rich middleware it eliminates ad hoc and temporary integration solutions. Ultimately it helps maintain a coherent, scalable and adaptive environment to handle present and future requirements.
One of the questions that comes up when looking at middleware is whether a business is better to build their own integration or purchase a middleware package. Believe me, when I tell you, middleware is the right way to go. With a custom integration stack and no middleware you are asking for trouble. Not only is building a custom integration time consuming and expensive but you also have to consider what happens if data is lost, or the system fails. Do you have a contigency plan or an audit trail? As I said above, middleware allows for greater control and flexibility with your data. Still not convinced that it is better to buy then build, then take a look at the infographic and go through the questions we pose.
Learn More About DIY and Custom Data Integration in this Infographic
Like this article? Share it below.