24 April 2008

To smooth its micropayments process for deposit refund, supermarket chain Netto got a new IT system that enabled it to define and implement processes to resolve data errors and thus improve cash flow in the business.

In Germany, an amendment to packaging laws is giving retailers financial headaches. While retailers are obliged to simply pay customers when they hand back all kinds of cans and bottles for recycling, there is a complex procedure for retailers to resolve recompense from the manufacturers – the so-called ‘putters-in’ to the system. Retailers pay a mere 25 cents per item, but multiplied by millions of bottles, this process results in a lot of money outstanding.

An expert at Deutsche Pfandgesellschaft (or DPG, the organisation that oversees the system of returning bottles and cans) has estimated: “More than €300 million of deposits on non-returnable beverage packages could not be resolved between German retailers and bottlers within the first three months after the third amendment of the German Packaging Ordinance became law.”

Problems were caused not so much by the physical taking back of the packaging units. That was mainly automated by reverse vending machines installed in the stores. But, as the packaging units themselves are collected and destroyed in a separate loop, only the very small data records produced and signed by the reverse vending machines remain for clearing the payouts to the retailer from the manufacturer (the company that filled the bottle or can).

The legal definitions for this task do present some stumbling blocks. The raw data must be collected from the machines in the stores and firstly verified signature-wise and also realtime content-wise with the DPG master database. Also, the originator must be identified and the data collected must be split up for every consignee, aggregated and transformed into a valid claim. For finally resolving the claim, all market participants need to be able to communicate electronically with one another. Different communication standards for six communication paths were defined. Flat file, XML and EANCOM message types with newly worked out data fields were allowed to exchange the data and AS/2 and X.400 were specified as communication protocols. In addition, the signed raw data must be forwarded 1:1 with every claim to allow the deposit account manager to perform plausibility checks. And finally – and here comes the main cause of delays in resolving payments – it was agreed that the entire claim could be rejected if only one single data item wasn’t correct.

One single mistake in the complex procedure of mass data processing – a data record representing a bottle worth 25 cents not correctly processed, the unit not correctly identified because of a synchronisation problem with the DPG master data base, the wrong consignee assigned for it or a problem with the signature validation – and a claim worth millions cannot be paid until the problem is solved, the data corrected (in the worst case manually) and the whole procedure run again.

Therefore, not surprisingly, Netto wanted to be fully in control of the process and its data. It decided to let Quibiq – its Microsoft Gold-Certified partner for enterprise application integration (EAI) and business process management (BPM) solutions – implement a specifically customized solution to handle the complete process on its existing Microsoft EAI platform.

Already in 2005 Netto had decided to use Microsoft BizTalk Server as its central EAI platform. In 2006 Quibiq implemented the complete process of daily supply ordering for more than 1,100 branch stores via seven logistics centres across the whole of Germany on this platform. That is now the basis of Netto’s ability to react flexibly and efficiently to new market demands or compliance issues like the Packaging Ordinance. The deposit clearing solution was implemented as another process on the existing platform. In this case, it was also upgraded during the implementation of the most recent edition of Microsoft BizTalk Server – 2006 R2.

The EAI platform was, from the very beginning, designed to allow the flexible implementation of various kinds of specifically-defined business processes, as well as to securely handle and exchange mass data. The platform has enabled the project to meet its deadlines, has kept the budget to a reasonable size and put Netto fully in control of it. BizTalk also provides advanced business intelligence features like business activity monitoring and real-time analytics, so every step in the process and therefore every possible cause for a processing mistake, is fully transparent to Netto.

The focus of the project was not so much the clear-cut run-through process as it was laid down legally by the DPG. The error handling was the main issue, to allow the operations team to react quickly if, for example, invalid bottles are detected at a specific store, which helps in protection from fraud.

During workshop days, the team spent a lot of time on the whiteboard and that time was well worth it as the complete process and all its possible contingencies were commonly defined. Because BizTalk comes with a graphical process design tool (with standard interfaces to VISIO or ARIS) the implementation was then not a question of writing thousands of lines of code but rather to simply depict what was defined by the process experts.

With the powerful, flexible and secure EAI platform in place, Netto is confident it can handle all the new tasks the marketplace, the competition or the legislator may demand of it – today and tomorrow.

This case study first appeared in the Spring 2008 edition of Retailspeak magazine.