“Should we automate?”--It’s not a question we often hear at Mimica.  

What should we automate?”--That’s the prevailing question of today. 

It’s a question posed by leaders regardless of where they are in the automation journey:  

  • New to automation and not sure where to look 
  • Have had success with automation but not at scale 
  • Have successfully automated “low-hanging fruit” and are looking for what’s next

It’s understood that automation drives business value. The challenge is identifying where to focus to realize that value. Scaling automation depends on identifying and developing the highest ROI opportunities. In this post, we look at different discovery methods for RPA and other types of automation.

Manual Discovery: Top-Down (Interviews, Workshops, Observation)

This method is often used when an automation program is just starting out. An example of how this plays out might look like this: A finance leader determines that automation is a key initiative for their area of the business. They employ the help of automation experts to go into various functional areas and find the best automation opportunities. Armed with knowledge of where automation opportunities are commonly found within finance (Accounts Payable, Order Processing, Reporting, etc.) the automation team holds targeted interviews and workshops with key players to identify manual, repetitive work. 

This can be an effective method for finding the most apparent opportunities, but not so much after the low-hanging fruit has been automated. Teams quickly realize that this is not efficient and limited in effectiveness. It is also highly manual and resource intensive.

Manual Discovery: Bottom-Up (SME Sourced) 

An alternative to deploying automation experts to find opportunities is training business teams to find them on their own. There are two main components that are needed for this to work: Education and Leadership Support. 

Educating business teams on what makes a good automation opportunity is critical for this approach. An effective way to do this is through internal roadshows. Once a team has a few successful automation projects under their belt, they can evangelize the results to other teams throughout the organization. This is a great place to start, and as automation governance begins to mature education can become more formalized.

The second key to success with this approach is leadership support. When automation is a strategic priority, employees can play a direct role in hitting organization-wide goals. Without buy-in from leadership, identifying opportunities can feel like just another task.

A potential limitation with a bottoms-up approach is that people often don’t know what happens outside of their role, so end-to-end opportunities tend to be missed. Another limitation is that even after training, employees don’t realize how much of their work is actually automatable. For example, a process may not be automatable with RPA in its current state, but could be through leveraging an alternate data source. This leads to missed automation potential. 

Automated Discovery: Task Mining

While manual methods for identifying automation opportunities can be effective at first, they ultimately leave massive automation potential on the table. Task mining provides an objective, data-driven approach to process discovery. This not only means that more opportunities are uncovered, but also that the opportunities lead to higher ROI. It’s good to know that a task is automatable, but it’s even better to know how difficult it will be to automate and what the impact will be once automated. These are exactly the types of insights that Task Mining provides. 

Task Mining works by collecting data from the clicks and keystrokes that users perform on their computers. This means that any time people are working on a computer, task mining can be used to generate insights.The number one use case for task mining is accelerating automation deployment. Here are some of the reasons to consider task mining over manual process discovery:

  • Task Mining discovers all automation opportunities: manual discovery is limited by the knowledge that people already have of their processes. Task mining continuously observes SMEs working on their computers, ensuring that all automation opportunities are identified.  
  • Task Mining uses a data-driven approach: task mining uses real process data to calculate metrics on automatability, complexity, and impact. This enables better prioritization than manual estimates that are often subjective. 
  • Task Mining is efficient: speed and scale are two of the main reasons to pursue task mining. Opportunities can be identified and mapped in as little as a week. Task mining also requires a fraction of the resources that manual discovery requires. 

Task mining is the most effective way to identify automation opportunities at scale, but that doesn’t mean that traditional methods should be ignored. One of the keys to manual top-down discovery is focusing on areas with a high-volume of manual work. This is also a great way to approach task mining. Another critical element of traditional discovery–leadership support–is equally important when deploying task mining. The most successful automation programs combine task mining with elements from traditional process discovery.