In 2009 we conducted a study on the inter-relationships of B2B e-commerce deployments and Enterprise Resource Planning (ERP) projects. One of the issues we explored was the level of exception processing required for supply chain transactions that flow directly from a B2B gateway into an ERP system. AMR’s study found that on average 2.9% of the electronic transactions required some manual intervention. I was surprised to find an almost 3% error rate on average, especially for a highly automated process such as supply chain integration. Assuming that each error costs $30 to correct, a $5B manufacturing company could be spending $500K per month just for exception processing of supply chain transactions.
What Types of Transactions Cause Exceptions?
The supply chain transactions that I am referring to are, of course, documents such as Purchase Orders (POs), PO changes, advanced shipment notices, bills of lading, commercial invoices and remittance advices. Over 1/3 of data in a typical manufacturer’s ERP system originates from trading partners such as customers, financial institutions, logistics providers, contract manufacturers, distributor/wholesalers and direct materials suppliers.
Supplier data is notoriously bad, especially from smaller vendors. Common examples of bad data passed from suppliers into ERP applications include:
- Suppliers over-ship or under-ship relative to quantity requested
- Invoices raised by suppliers with no corresponding PO
- Aggregating items from different POs onto a single invoice
Logistics data is frequently incomplete and inaccurate due to missing fields or different uses of codes within EDI transactions. Examples of bad data include:
- Shipment status messages missing date and timestamp for an activity
- Quantity populated in shipment message, but without unit of measurement
- Transportation documents arriving out of sequence
Customers are the most challenging trading partner in terms of data quality. They are the greatest source of bad data, but often are the group that companies have the least amount of influence over. Examples include:
- Orders for discontinued SKUs, GTINs or Part Numbers
- Requests for ship-to-locations that are not registered with supplier
- Out of sequence purchase order changes
Impacts of Exceptions
Let’s examine the 3% exception rate and its impacts on the supply chain in greater depth. Many companies experienced much lower exception rates than 3%. For example, 20% of respondents indicated a lower than 1% fall out from supply chain transactions originating externally. However, there was another group that performs far worse today. 20% of these respondents indicated that greater than 5% of B2B documents required exception processing. A third way to view the data would be to state that 79% of companies experience a greater than 1% exception rate.
I mentioned earlier that a $5B manufacturing company could be spending $500K per month on manual processing of supply chain transactions that were intended to be automated. The monthly spend is highly dependent upon the monthly transaction volumes of a manufacturer. Below are a few examples of the volumes exchanged by a few of our larger customers:
- German-based consumer products manufacturer – 300K per month with 15% growth
- US-Based food manufacturer – 700K per month with 28% growth
- Swiss-based steel manufacturer – 150K per month
- French-based entertainment products brand – Growing at 40% to 1M per month
If we assume the cost to manually resolve a failed supply chain transaction is $30, which is conservative, then we can arrive at a graph such as the one below charting monthly costs from exception processing.
I will share more thoughts on the implications soon.
