In my last blog, the 2nd in our 3 part series on the “Evolution of the Modern CoE” I discussed the critical tools that are required to control and optimize your digital transformation. As you may recall, these included process definition & assessment, automation, change management, collaboration, model-based reporting and monitoring. With so many tools available, it’s crucial to understand how and when to use them. In this post I will dig deeper and explain the methodologies necessary to successfully apply your toolkit.
As you read through the methodology, keep in mind that you may choose to use different steps at the appropriate times depending on what stage of bot adoption you are in. You can dive into any particular step that makes sense based on where your organization is in the bot adoption process.
Assuming you don’t already have widespread bot adoption, the first step in the methodology is to identify which process to start with. This is where the inventory of your processes comes into play; start here and identify the specific processes you’re most interested in digitally transforming. You may want to go through a multilevel prioritization process, which means that you’ll want to prioritize your processes based on your strategic initiatives. Let’s take the customer journey as an example. Each touch point in the journey has a relationship to an actual process, and as you identify issues, you identify the processes that are the primary candidates for digital transformation.
Then you’ll want to assess the process candidates you’ve identified and describe all the tasks underneath them. Once you’ve completed this activity, you may want to look at the steps and the processes in the context of risk mitigation or capability gaps, etc. In addition, you may want to look at high leverage control processes through some of your organization’s biggest risks.
After you’ve identified and prioritized your processes, the next step is to make an assessment for automation and RPA maturity. There should be a standardized approach to determining which key tasks are best for RPA. This assessment is critical, because otherwise you might fall into the trap of assuming that everything is “RPA-able” – and it’s not. In fact, one of the keys to transformation success is understanding which tasks are suited for RPA are and which are not. For example, repeatable processes are highly RPA-able, but on the flip side there are plenty of processes that have many exceptions and are not ideal candidates. In the case of RPA, you always want to create the highest impact. This also means you may need to do a process redesign along the way in order to ensure you’re not implementing and encoding a bad process to begin with.
Now that you’ve identified, prioritized and assessed your process, you need to validate the steps and business logic associated with it. This will help you uncover the places where the bots will go. You’ll likely start by referring to your process diagrams or models, which tell you what you *think* you do. Next, you can apply process mining to help you validate those assumptions. To do that, you can compare and contrast the mined model to the assumed model. With the mined model where you’ve done the validations, you can actually use the mining tool to do some analysis to validate the best opportunities for the actual points where the bots may go. Essentially, you’re identifying the bot objects in the models and then linking that to your compliance.
In cases where you are at a high level of bot adoption already, this is the point in the methodology where you may wish to begin. Either way, this is the point where bot objects are created in the model.
This is also a good time to identify the systems or applications the bot will use. By identifying them, it allows you to understand the impact of change to that system or applications on the bots that are dependent upon them.
Once you understand your process and where your bots are going to go, it’s time to identify risks. For example, if your bot is going to touch data, data protection will play an important role. You can identify and tie it to your overall compliance issues. You should leverage the model and the relationship to the processing task, and then create the relationship so that you can report on it and understand what’s truly taking place. You should then tie the bot governance directly to the business governance; look at the mitigation strategy to ensure the controls you have put in place on that bot in that instance are in alignment, and importantly – stay in alignment. Key metrics need to be identified as well. You want to be able to watch the bot or collection of bots or the process itself to make sure that all stay in compliance.
After the steps above, you are finally to the implementation step. This is where service partners can become very valuable in terms of execution. Each execution will be different, but the key to remember here is to monitor access points where data can be extracted and looked at in the context of the dashboard.
The last piece of the methodology is monitoring. Now that you understand the metrics you should be looking for: 1) tie those metrics to a dashboard, 2) track risk control effectivity, and 3) monitor the limits of the bots. If the bots go outside of your governance limits, you can drill down, identify who owns them and understand how to deal with the challenge. The key is having the model in place for easy identification.
Digital transformation is no easy task but carries with it the ability to be an incredible catalyst for companies of all sizes. To help, we have now explained the elements of digital transformation, and given you both the critical tools and the essential methodology needed be successful. If you have further questions, or need more guidance, you can also contact one of our transformation experts here.