Not knowing what the overall goal is can easily ruin your business optimisation benefits. The problem is that if you don’t know the overall goal you might end up focusing on the wrong elements in the optimisation process.
For example; a simple process for buying lunch vouchers in a school canteen could look like the process illustrated below. Students buy a voucher, queue-up for lunch, hand-over the voucher and get their food in return.
The process has some challenges:
The process time for process 1.0 and 1.1 is high. One problem is that a lot of students stands in queue and tries to buy a voucher (process 1.1) only to find out that they don’t have enough money – then, they either decide not to buy lunch (ends process) or they go back for more money and re-enters the queue. In both situations the processing time for process 1.0 and 1.1 is high due to inefficient process.
Another challenge is when the students waits in line in the canteen to get their food. Here the processing time is high (in process 1.4). One of the problems is that a lot of students queue-up without having a valid voucher. It takes time to clear that out. The students being rejected might return to process 1.0 in order to buy a valid voucher (not diagrammed).
All right, lets imagine the process must be optimised. What would you change in order to create a more optimal process?
One suggestion could be the optimised process illustrated below.
In this optimised process one of the canteen guys now stands in the doorway to the canteen and acts as a gatekeeper (process 1.8). Before a student is allowed access to the food queue he checks that the student is having a valid voucher. This optimises the overall processing time for the 1.4 process since there is no processing time going to students with an invalid voucher.
So, is the process now optimised the way we wanted? It depends on what goal we are trying to achieve. Lets have a look on some potential business goals:
1. The business problem we are trying to solve is that the space reserved for queuing-up for buying vouchers is limited and sometimes students give-up buying vouchers (canteen is loosing customers). The business problem to solve is to process buying vouchers faster in order to increase capacity.
Was this business goal achieved? Not really – the optimised process is process 1.4. Maybe it would have been better to look at ways to reduce process recurrence. Why are a significant number of students not able to complete the process of buying vouchers without having to cancel the process or go back for more money? Don’t we have signs informing the students about prices? Do we only accept coins? Do we only accept exact amounts (not giving change back)? Etc. Do we only allow buying vouchers for the current day and not for e.g. the entire week? Could we eliminate vouchers and let students pay with cash when getting their food.
2. The business problem we are trying to solve is that the capacity in the canteen in very limited and there is not space enough for the queue in process 1.4. Process 1.4 needs to be performed quicker.
Is this goal achieved? Actually it is (assuming we know the processing time and numbers) because we have introduced a gatekeeper to eliminate processing students with invalid vouchers. But wait; maybe executing the entire processing in the canteen swimlane could optimise the capacity.
In the diagram below finding food (process 1.5) has been parallelised so that it happens simultaneously with process 1.4 (maybe the order when buying the voucher is electronically transferred to the canteen).
Notice: Maybe the business problem to be solved is that the total processing time should be optimised because more students should be able to use the canteen and as fast as possible get to process 1.7. This would require optimising the entire process and would involve all examples used – this has not been diagrammed in this post.
The lesson learned here is that it is crucial to know the overall business goal in order to design the optimal to-be process.
When optimising a process look for the following:
- Are there any (sequential) processes that could be executed in parallel?
- Are there any processes that could be eliminated?
- Are there any manual processes that could be automated?
- Are there any decision shapes directing the process in alternative routes that could be eliminated or reduced?