BPMS projects are often under immense pressure to deliver quick business value and benefits. One of the key questions organizations implementing BPMS often ask is, “Which process do we implement first?”. The answer to this question is extremely important since the first process can set the pace for the entire BPM Project.
Thomas Stoesser in his article, 5 Questions to Ask Before You Automate Your Business Processes, outlines some interesting questions to ask in identifying which process would be most suitable for automation.
One effective way to shortlist processes is to examine existing business processes from a waste reduction perspective. There are 7 types of waste that can exist in any manufacturing or service process. In identifying the first process to automate, you could for instance, shortlist processes with any of the following waste elements:
1. Overproduction: Shortlist processes that generate more output than required or processes with operations that continue long after they should have stopped. An example is printing paperwork that may not be needed or processing scheduled orders regardless of current situation/changes in demand. Overproduction increases inventory costs, waiting time and unnecessary motion.
2. Waiting: Identify processes with extended periods of inactivity. This can happen when a preceding task takes too long to deliver the output needed to start or complete the next task or a process participant is unavailable, leading to the creation of a backlog. Examine the processes in your organization to identify the ones that involve idle waiting time for possible improvement.
3. Defects : Defects leading to scrap or rework can indicate that a process needs to be improved. Beyond manufacturing/physical defects, examine customer complaint files containing issues and requests for correction to shortlist potential processes that can be improved. An example of a defect is a data entry error.
4. Over-processing: Examples of over-processing include entering the same data into multiple systems, making multiple copies of documents and generating reports that are of little or no use to anyone. Identify processes with any of these characteristics for possible improvement.
After shortlisting, the next challenge is identifying which process should come first.
I'll now share a BPMS case study with you.
During the initial phase of our Business Process Management Project, the team was faced with the dilemma of deciding which process to automate first for the BPMS pilot implementation; we needed to show we could deliver business value quickly with the BPMS. After some research, we discovered that the general recommendation by BPM pundits is to start with HR processes, since they are easier to manage and are less risky.
We did not, however, want to be guided by this, so we identified some criteria for selecting the first process as follows:
- The process had to be simple, with minimal conditional routing, to avoid delays associated with analyzing and designing a process with complex conditions.
- We wanted the first process to be accessible by a sizeable number of users so that the benefits could be experienced by as many employees as possible.
- We did not want the process to span too many departments, to reduce the complexity of training, issue resolution and general operations management.
- We wanted a process that had major opportunities for cost reduction; a major organizational goal at the time.
- We wanted the process owner and other relevant stakeholders to be excited about the BPM project, so we considered their attitude toward the project.
- We wanted a process that would fit in with the existing BPMS feature set, to reduce development/customization time.
- We wanted a process with no external users like clients and suppliers, etc. It had to be strictly internal.
Taking all these factors into consideration wasn't an easy mission. We made a lot of assumptions along the way; some turned out to be correct while others were not so correct. We shortlisted a few internal processes based on the above criteria and eventually selected the first process for automation. We arrived at the same conclusion as our initial research: an HR process scored the highest, using the weighted scoring model and became our first process.
A bigger surprise came three months into the project. The process we selected was not as simple as we had thought it would be. We had already announced the first process to be automated to stakeholders and had started analysis before we discovered just how COMPLEX it was. Yes, the process was simple, but not the business rules that govern it.
Selecting the first process for automation should be done only after complete analysis. A superficial analysis of a process, without considering the business rules governing it, will not yield sufficient information to determine just how complex or simple the process is.
When analysing a process, always consider the number and complexity of its business rules, the location of process participants, exceptions and the availability of clean process data.
Regardless of the process you choose, remember to keep your eyes on the prize: value.