Automating repetitive tasks, automated approvals, and scheduled report generation are among the top things companies automate. Businesses make large investments to keep up with the trend. Most businesses successfully implement these automations, but few maximize the return on their investments. Here are the key factors to maximize your automation investments.
Business Process OptimizationMany automation discussions begin with a process map and use case discovery to identify areas where automation can excel. These opportunities are then ranked by the number of human capital hours saved or the volume of the tasks. The solutioning and implementation phases follow.
Technology is designed to improve how we work, so automating existing business processes is a natural fit. At a minimum, automation should improve the speed of the processes, as technology can process information and perform calculations much faster than we can. However, before getting into the details of what and how to automate, it is important to ask why.
- Why does the process exist?
Determine the value it produces. If it does not appear to add any value, it probably does not. As long as you have conducted thorough research from all perspectives, do not be afraid to be a trailblazer for your company and suggest removing unnecessary procedures.
Example:
If a process calls for approval from multiple levels of approvers, and the approvers make their decisions based on a defined set of rules, then the approval logic can be fully automated. Otherwise, the goal of approval is to create a record of acknowledgement of the decision. As long as we accomplish the same goal, the mandatory approval process may or may not be necessary.
Once we have a solid grasp of the process, consider if there is a more efficient way to complete it.
Example:
Users who want access to a system must complete specific training courses. The existing process goes like
Alternatively, what if we publish training information and course requirements? Then the new process can be
- Why does it have to be done the way it is done now?
In the context of automation, there are times when people want to include human touchpoints to make sure the process is functioning as intended. However, the truth is that human touchpoints introduce hazards to an automated process. Circumstances that can be disastrous are generally unavoidable regardless of whether or not human touchpoints exist. That is not to say that we should never have to be concerned about fully automated processes. A well-designed audit process can achieve the same goal without introducing hazards to a fully automated process. Do not hinder your automation initiative from producing the best possible outcome.
Data Alignment
Data is the foundation of automation. No matter what you are automating, there must be some common fields that allow the systems to connect to each other. Ideally, the flow of data should be defined. If not, the data may become ambiguous, which will increase the complexity of the automation. Data transformation can be completed through automation, but it should be the last resort. Ideally, your company's data governance should ensure that the data for the common fields all follow the same standard.
Example:
In the access request example above, the common field that connects the learning system, the request system, and the destination system of the access request is likely an employee ID field. Passing the employee ID field from one system to the next as the access request goes through the design workflow will ensure that the data is unambiguous. This is because it is the only way that downstream systems can obtain a new employee ID for any given employee.
However, if one of the downstream systems pulls the employee data, including the ID field, from the HR system, the data becomes ambiguous. This is because we do not know whether the employee has gone through the access request process. We must find a way to identify which employees are approved to have access to the system and then assign the appropriate access to them in order to support the access request automation properly.
It is not always necessary to have unambiguous data. However, if we have the choice, we should take the opportunity to keep the automation simple and straightforward. Finally, if the employee ID in all three systems are stored in a different format, we will have to handle the data transformation through automation. Ideally, all of them should be the same so that we never have to guess or go through more steps to identify the correct employee.
