Before you can effect any Process Improvement you need to have a clear, accurate and reliable picture of the AS-IS condition of a Business Process in terms of its Key Performance Indicators (KPIs). These KPIs can be measured against SLAs if it is outsourced, or in the Shared Services context. In case it is not outsourced, companies may need to benchmark their KPIs against the Industry Best. There is hardly any point in improving a Mortgage Loan Processing process from 21 days to 20 days when your competitor has an online system and they do the same process in 5 days! Sometimes you may need to leapfrog over your own old benchmarks! Incremental process improvement may not help you face the competition effectively, in the marketplace!
This article Business Intelligence Adds to Process Reengineering talks about this very eloquently. The authors make the point that relevant KPI performance data needs to be collected consistently all the time.
They make the point that sometimes all this data may be spread in many places and need to be collected in one single set of data tables using Extract, Transform and Load (ETL) functions. Analysis in terms of dashboards and alerts can then be of use in process improvement.
This is the classic problem many business processes have. They are all run end-to-end with multiple heterogeneous software systems in the background. If the process is outsourced or parts of the process is outsourced, then data gets dispersed between multiple companies. This makes it very hard to assemble ALL of the relevant KPI data in one place.
For example, let’s say in an Order to Cash process, the collections part of the process is outsourced to an external vendor. This company has all the data upstream and downstream, but in between for the collections process, the data is in the vendors’ systems somwhere else.
This is where measurement is usually done with whatever is available, rather than what is needed for meaningful analysis and improvement!
Business process improvement is not possible without the collection of the right data at the right time and making it available in the right form to the right people. Easier said than done! Legacy considerations and history of processes within companies still support a Functional View of operations (Finance, Marketing, Manufacturing, etc) rather than a Process View. WIth increasing automation and shortening process cycles, this is becoming somewhat easier!
The best performance improvement is the transition from the non-working state to the working state – Anonymous
