As organizations around the world navigate the global COVID-19 pandemic, business continuity and remote work, once considered long-term needs, have become critical to their survival. Digital transformation is evolving from an overhyped buzz phrase to an actionable goal for them if they want to remain competitive in the AI era. For businesses that struggle with driving digital transformation, starting with a Robotic Process Automation (RPA) project is a smart move.
According to Gartner, the RPA market grew 63% in 2019, making it the fastest growing in the enterprise software category. This comes as no surprise. Compared to data science and artificial intelligence (AI) projects, RPA project costs tend to be lower, the technical skills less demanding and the change management process is much easier. Although the technology is not as mature in terms of security and scalability as other enterprise software, progress is being made as an increased number of businesses adopt the technology to optimize internal processes.
RPA is transforming the employee experience by automating mundane tasks and allowing employees to focus on more intellectually challenging and value-adding activities. Happy and productive employees in turn focus their attention to providing excellent customer service and improving the customer experience. The first step toward a happier, more productive workforce is identifying the right automation opportunities.
Identifying Good Automation Opportunities
RPA is best suited to automate tedious and repetitive manual tasks. To come up with a list of good RPA opportunities in your organization, walk around your operations department and take a poll: What is the most boring, tedious, time consuming manual task that has to be done on a weekly basis and could probably be taught to a five year old? In a few minutes you will likely end up with a list of tasks like the following:
- Download X report from Y website, export it to Excel, clean it up and format it, and email it to the management team
- Copy data from the new snazzy X web-based application to the old legacy dinosaur system
- Key-in information sent by customers via email to the claims system, upload the email as an attachment
- Extract public data from PDFs published on government websites
- Reconcile financial data between reports from X, Y, Z departments
Any of these tasks would be a great choice for an RPA project. Activities that involve formatting files, scraping websites, extracting information from documents, reconciling data and emailing groups of people are among the most common use-cases for RPA automation. That is why businesses in banking, financial services and insurance sectors are increasingly adopting RPA bots to interact with their legacy systems, extract data from multiple sources, automate workflows, create consistent reports, and update data in real-time.
A common mistake in RPA implementations is to pick a use case that is seemingly simple but too complex in practice. According to Forrester, you should apply the rule of fives to evaluate whether a task is a right fit for RPA. The task should involve fewer than five decisions, touch no more than five applications and be completed in 500 clicks or less. Once you have established the RPA automation goals and activities, it is important to set up the project for short and long-term success.
Leading Successful RPA Projects
Compared to enterprise software development projects, RPA bots can be built in much less time and have tangible results and business impact from the very first deployment.
- No migration– RPA bots interact with legacy systems; they do not replace them. An RPA implementation does not require migrating data or business functions to new applications, which takes a significant amount of effort when upgrading legacy systems.
- Minimal coding– With enterprise RPA platforms like Automation Anywhere or UIPath, automating a task requires minimal code, and often no code at all. Resources from many different backgrounds can be trained to configure bots; no programming experience is necessary.
- Buy-in from the end user– End users sometimes resist the changes in processes that are introduced by new software implementations. However, when a user sees an RPA bot complete a task in minutes, when it usually takes them hours, they are amazed and relieved that a bot will be able to relieve their burden and make their lives easier.
Building a bot that can scrape a website, reconcile a report or send an email can be quite simple. Complexity will increase as the number of steps, applications, and business scenarios increase. Demoing a simple automation is an effective way to get buy-in from the end users and gain momentum for the RPA project. However, it is important not to fall into the trap of continuing down a path of accumulating quick wins. Bots will be doing work that people used to do, and the organization must take responsibility for their performance, continuous improvement and security.
Building the organizational structure is one of the most challenging aspects of an RPA implementation and the most critical for its long-term success. Who reviews the output produced by the RPA bots? Who updates it if the underlying data, processes or functions change? Who is responsible if the bot fails? To answer these questions the organizations must establish a framework with the following key components:
- Oversight and governance– Like employees, RPA bots need to be included in the organizational structure. They should have a direct supervisor who oversees their daily tasks and coordinates updates and maintenance with the rest of the team. If the bot encounters an unexpected error that prevents it from completing a task, the supervisor should always be kept in the loop. A hierarchical structure of responsibilities, communication and accountability must be clearly defined in an RPA project.
- Maintenance– Most RPA bots need maintenance when data changes, webpages are updated, or workflows change. Like most applications, they require both scheduled and unscheduled maintenance beyond the RPA project. Many organizations create RPA Centers of Excellence where interdisciplinary teams are trained to maintain bots.
- Security– Proper security measures need to be taken when implementing RPA in highly regulated industries like Financial Services and Healthcare. If bots are dealing with sensitive information, data should be encrypted, bots should run in a secure hosting environment, and access to the bot itself should be limited.
The Future of Intelligent Process Automation
Organizations taking advantage of Robotic Process Automation can extend their capabilities with the advances in Intelligent Process Automation (IPA). Beyond automating the manual, repetitive tasks, IPA shows promise with augmented capabilities powered by artificial intelligence (AI). Here are some use cases that are being researched and developed in IPA for future use in the industry:
- Process optimization- An AI model may suggest ways to optimize current processes.
- Natural Language Processing- An RPA bot may become more human-like and engage in more Slack and email communications with employees, detecting intent and responding with natural language.
- Predict system maintenance– Artificial Intelligence-powered bots may automatically detect changes in data and websites and sometimes even automatically correct the code.
Digital transformation will continue to be a popular topic as businesses create services and products powered by technology that provides compelling customer experiences. RPA projects provide a great opportunity to obtain quick wins and long-term impact around process optimization and employee satisfaction, and paves the road for future intelligent automation.