Recently Margo and I have visited an ISV who produces software solutions for the retail industry. In one of the meetings I learned that retail is a business of low margins and extreme competition. This is specifically true for countries like Germany. As if I asked for an example of this claim I heard in the news that a very large US superstore chain would retreat from the german retail market in order to concentrate forces into markets where they can grow faster.
So when margins are low, prices cannot be raised, costs must be kept down. The software that these folks write is to harmonize the demand and supply in a way that customers would never stand in front of empty shelves while at the same time, the store's stock is never full. In the past, this optimisation was part of the staff's responsibility and based on vast experience, however with the variety of products these days and the dynamicity of the labour market it is impossible to build the keen sense for quick and slow selling items. Also, more and more stores are branches of a chain where smart logistics is to be applied. Smart software is required here to forecast demand.
One of the most complex parts of managing a supply chain is the integration of the various parties in this chain: stores, warehouses, manufacturers, shipping companies. To complicate things, goods and the corresponding data does not flow into one direction only. Flows are back and forth, e.g. if goods are returned due to quality problems.
For instance, one integration use case is when orders are sent from the store to the warehouse. Orders from all stores will be consolidated to create another order to the manufacturer. In some cases, the manufacturer's shipping company sends the goods directly to the store, in some cases these products are sent to the warehouse for redistribution.
Sounds like a big integration task!
Another example is the creation of orders. Orders are based on forecasts. Forecasts are based on existing sales data. The sales data needs to be transferred from the stores to a central location for calculating the forecast and the orders. The current solution that transfers sales data to the central server is based on a simple yet proprietary protocol. It works no problem. However their customers need to operate and support this proprietary protocol and have no real alternative. The ISV considers using an Enterprise Service Bus to open up the architecture of their product to use any protocol.
Now what does this have to do with Open Source?
The ISV has OEM licensed its product to a large SCM software company. The inclusion of commercial third-party products into their product "could" mean that existing contracts need to be renegotiated. Hence, this ISV envisages open source software as a strategy for creating value for customers, Partners and themselves without having to change business processes.
Another note I wanted to add is that, the described use case with the transfer of sales data is a challenging one for us and the Celtix product. The integration we're seeing here is not an integration of functionality, i.e. classical application integration. We're looking at integration of data sources and sinks. To my mind, there are many such integration problems out there. We've just seen a tiny bit of it.