Automation Academy

Guide to Automation: Identifying the right processes to automate

25.06.2019
6 min read
Guide to Automation

Nowadays, a lot of organizations feel the need to automate internal business processes for the sake of cutting down operational costs and improving productivity.

Automating business processes in your organization can be incredibly beneficial, potentially allowing organizations to bring down the cost of automated job functions by 30-40 % while doubling their efficiency. However, automation gone sour can mean resources spent in vain and may discourage organizations from pursuing other automation projects for a long time, eventually putting them behind competitors who embrace automation and make it work.

The failure of an automation project is often attributed to a poor choice of a use case for the initial trial or proof of concept (PoC). According to some research, that’s possibly the reason for more than a third of unsuccessful automation implementations. 

Identifying the correct automation opportunity can be a challenge, but it shouldn’t discourage you from starting to implement automation in your organization. So, how to select the right processes for automation?

Step 1. Understand the difference between humans and computers

The first thing you need to do to select the best process to automate is to understand the differences between humans and machines.

Humans shine when it comes to dealing with uncertainty, ambiguity, and novelty — but machines are not yet able to handle them and often cannot navigate unexpected changes. This is also true when it comes to judgment or decision-making in such changing circumstances. Computers cannot fully replace humans in these conditions.

Humans are good at working with analog and unstructured data, whereas computers need digital and structured data.

Humans are less capable in consistent execution of the same processes. Repetition of routine tasks hinders concentration, resulting in errors. Computers, on the other hand, excel here. Also, tasks performed by computers can scale massively, which is not easily achieved with human workers.

These differences define the factors to be considered when selecting the right process for automation.

Step 2. Determine if the process can be automated

Based on the above, we can formulate several main factors in defining the process best suitable for being performed by a machine and where we can eliminate human errors.

Of course, selection of the best process depends on the tools you use to automate. A process not suitable for Robotic Process Automation (RPA) can be a perfect fit for Intelligent Automation (IA) augmented with Cognitive/Artificial Intelligence capabilities in addition to RPA robots.

How to choose a good RPA use case

Select a process that is rules-based. A process where employees follow a strict set of predefined rules is more suitable for RPA than one that includes decision-making open to employee judgment. About 70–80% of all jobs and key responsibilities in organizations are rules-based and thus can be automated with RPA tools. If the process does require judgment by the employee, try breaking it down into smaller parts, and you will see that most of them are rules-based and can be automated with RPA.

Make sure the process is routine and well-defined. The automated process needs to be established and shouldn’t change frequently. Tasks performed in the process should be the same, processed in the same order, and use systems that don’t change or rarely change. 

Consider the input and output data involved in the process. The ideal process will use structured data and readable electronic inputs. Examples of such inputs include:

  • spreadsheets
  • databases
  • JSON files
  • CSV files
  • XML and HTML files, etc.

Because of their digital and structured nature, they're a better fit for RPA than scanned documents, faxes, handwritten forms or emails.

For example, RPA can help successfully automate data entry processes, procurement, accounting and sales processes, etc.

How to select the best processes to automate with Intelligent Automation

Consider processes that use unstructured data. When it comes to Intelligent Automation, the criteria for selecting the use case are broader. While RPA works best with the data that is digital and structured, Intelligent Automation armed with Cognitive/Artificial Intelligence capabilities is able to process data that is non-digital and unstructured. Advanced image recognition and optical character recognition (OCR) capabilities are able to digitize the data, and machine learning can process unstructured information from scanned documents, images and other data inputs that are not a good fit for pure RPA.

Include judgment-based processes. Intelligent Automation solutions allow the inclusion of certain cognitive decisions, enabling automation of end-to-end processes with minimal requirement for manual effort.

How to select the best processes to automate

Step 3. Evaluate the use case potential

Defining the process as a good fit for automation is not enough for its successful implementation. After having selected the use case, ask yourself two questions: Does it make sense to automate it? Is there enough momentum for it? 

Does it make sense to automate it?

Not all use cases that can be automated are worth automating. Automation of business processes should have a positive impact on your organization that outweighs the resources spent – reduced costs, increased accuracy of processes, elimination of human errors, etc.

The processes that are most likely to have such effects are:

  • Time-consuming and time-critical. High transaction volumes hinder human work but don’t affect the performance of robots. 
  • Repetitive. Choose business processes that are performed on a regular basis. Automating such processes will have a bigger positive effect than the ones that are executed occasionally.
  • Prone to mistakes and lack quality. Eliminating human error in such processes will raise the quality of work and decrease the time spent on correcting these errors.
  • Include many disintegrated systems that use the same data. Using the same data across many different systems results in a lot of mistakes when handled by people, so such processes are well suited for robots.

Apart from the positive effects mentioned above, automating such processes will spare a considerable amount of time your employees usually spend on performing such tasks (and correcting mistakes) and will allow them to focus on more meaningful work that required less routine and more decision making.

Is there enough momentum for it?

Before implementing the use case, make sure there is support for it in your organization.

  • Advocate the benefits of automation with interested stakeholders. Before starting the project, identify the key stakeholders in the process and advocate the benefits of the process automation with them. Including them in the process will not only help to identify potential pain points, and ensure there will be enough assistance during the project implementation but also help ensure support from management. 
  • Get buy-in from upper management. Automation may require spending considerable time, human and financial resources (depending on the process, of course) before the positive effects of the automated process will be visible. Make sure the company’s management supports the implementation of automation and are aware of potential costs.

Neglecting these factors might result in a failed automation attempt even if the potential of process automation is very high.

Find more useful information about whether you need automation in your organization or how to identify the right business process for it in Automation Academy’s free Automation Essentials course.

Want to know more about Intelligent Automation?