The Gist
- Automating boring and repetitive tasks. Robotic process automation (RPA) uses software "bots" to automate routine business tasks, providing benefits in various industries.
- Prioritize security. Security should be a priority throughout the RPA design and development phases to prevent attacks and ensure effectiveness.
- Impact on employee satisfaction. While RPA can improve efficiency and effectiveness, it may not be suitable for complex processes or those dealing with unstructured data, and may impact employee satisfaction and performance.
What is Robotic Process Automation (RPA)?
With all the recent discussion about the AI-driven ChatGPT, Google Bard and Microsoft’s new AI-enhanced Bing, the concept of robotic process automation (RPA) brings to mind humanoid robots replacing human employees, but it’s actually much more benign (and beneficial) than that. RPA actually refers to software “bots” that are able to automate routine business tasks. Let's delve into what RPA does, how brands are using it and the challenges it poses.
The Origins of RPA: Screen Scraping and Workflow Automation
Gartner defines RPA as “a productivity tool that allows a user to configure one or more scripts (which some vendors refer to as ‘bots’) to activate specific keystrokes in an automated fashion.” RPA first began to be used in early 2000 and was comprised of a combination of artificial intelligence, screen scraping and workflow automation.
Screen Scraping
Screen scraping, which was initially used in the 1990s, refers to the action of using a computer program to copy data (i.e. scrape) from a website or application so that it can be used by other applications or websites. Typically, it was used to capture data from a legacy application so that it could be used in a newer application, such as transitioning from Lotus 123 to Microsoft Excel. A more recent example of scraping is Common Crawl, which is an open-source dataset that has been generated from a multitude of websites that have been scraped. Common Crawl regularly crawls the web to create free datasets that can be used for training AI applications. Because screen scraping is limited by its compatibility with existing systems, businesses turned to more adaptable and extensible technologies.
Workflow Automation
Workflow automation, which also began to be used in the 1990s, uses software applications to complete tasks and activities without the need for human input. Basic workflow automation has been in use for hundreds of years, and includes assembly lines and manufacturing processes, although initially, it was based on machines or mechanical devices, rather than software. Workflow automation is used in many industries today, including customer relationship management (CRM) software where it is used to automate form completion and customer communications.
The Combination
The combination of screen scraping and workflow automation powered the first RPA software, and in 2015 RPA began to incorporate AI technologies, which was when it really took off. Although on-premise RPA still commands the highest revenue share, as brands move to a cloud-based IT infrastructure, RPA as a service (RPAaaS) is growing in popularity. RPAaaS typically outsources RPA to a third-party vendor through the cloud, and is usually available as a fixed service or via a subscription model.
The use of RPA has continued to grow, and as the need to automate workflow increased during the COVID-19 pandemic, it became a key component that enabled many brands to remain operational. According to a report by Grand View Research, the global RPA market size was valued at USD 1.89 billion in 2021 and is expected to grow at a compound annual growth rate of 38.2% from 2022 to 2030.
Examples of modern RPA software include Blue Prism, Automation Anywhere and UiPath.
Exploring RPA Use Cases
Because RPA can be used across different applications, platforms and departments, there are many uses for the technology throughout various industries. Typical uses of RPA include:
- Extracting Data From Different Formats: Data is often stored in different formats including text, handwritten notes, spreadsheets and databases. By using optical character recognition (OCR) technology, RPA is able to aggregate and read data from different formats with high levels of accuracy.
- CRM Integration: Many CRMs today use RPA to store, sort and organize customer information such as contact information, purchase history, preferences and personal information. Additionally, RPA is used with CRMs for the entry of customer offers, case management, orders and shipment details, and the renewal of contracts.
- Automation of Customer Service Requests: Customers today expect swift responses to customer service inquiries. RPA can be used to sort queries and provide initial responses to customers, segregate queries into different categories, automate the rules-based processes that many call centers use and aggregate customer data.
- Banking and Finance Process Automation: RPA can be used to consolidate data from various sources into a single financial database and process invoices. Printed invoices can be automated using OCR.
- Sales Order Automation: RPA is able to automate sales order entries, invoicing, and the removal of duplicate or erroneous data, and can alleviate sales employees from having to perform time-consuming, repetitive tasks.
- Data Scraping: Falling back to one of the original uses of RPA, the technology is still used to extract data and content from websites, allowing the data to be analyzed and used by other applications.
- Employee Onboarding: RPA can be used to automate many of the onboarding processes for new employees. It can be used to enroll employees, schedule meetings and classes, and send the new employees automated messages detailing their responsibilities.
Jan Arendtsz, founder and CEO at Celigo, an integration platform as a service (iPaaS) provider, told CMSWire that because RPA is a no-code application, business users are able to create automations. RPA doesn’t require coding or programming knowledge. Most RPA solutions feature a drag-and-drop user interface, which means that "RPA automations are generally implemented in days (or even hours in the case of the simplest of processes)," said Arendtsz.
Security Considerations for RPA
As with any business software, security considerations must be first and foremost. Jacob Ansari, PCI practice leader at Mazars, a global cybersecurity consulting firm, told CMSWire that although RPA can be useful for cost reduction, minimizing human error or otherwise doing away with tasks unsuited to human activity, the existing security risks in computing remain present in RPA and the security practices needed to mitigate those risks apply similarly.
Updates are often required as patches are developed to protect against security vulnerabilities, but those same updates can break the RPA process. This needs to be addressed long before implementation. “RPA efforts need security in the design, engineering and development efforts, particularly as it pertains to designing effective access controls, credential management and making use of a software stack that can be updated as necessary,” suggested Ansari.
Additionally, there are other considerations that must be incorporated into the RPA process in order to limit security risks. “RPA processes should get specific user accounts with the least privilege necessary to perform the job and that are locked against interactive use,” said Ansari. “Further, those credentials should be rotated regularly.” Ansari suggested that RPA processes should have logging and monitoring, and event detection and response efforts should be informed of these so they can look for anomalous or problematic behaviors and respond in an effective manner.
Security must be a priority throughout the RPA design and development phases. “The sooner these practices get integrated into RPA efforts, the more likely they are to prevent attacks and the easier they will be to institute. Trying to bolt on security after the fact has always come with additional costs and less effectiveness, and RPA scenarios are no different,” said Ansari.
Learning Opportunities
Webinar
Apr
5
Adopt a Modern Digital-First Marketing Strategy
Learn how to use a composable DXP to your advantage
Webinar
Apr
13
Composable Commerce The Future of Retail
The Adobe Commerce approach to composable and why it's best in class
Webinar
Apr
13
Accelerating Content Velocity Across the Digital Supply Chain
Learn to integrate DAM with upstream tools to improve workflow efficiency.
Webinar
Apr
18
Drive Business Growth Through Customer Service and Support
Defining the future of technology. What is AI’s role in the future of customer service?
The Challenges of RPA
While RPA can improve the efficiency and effectiveness of many processes, there are many cases where RPA is not an effective solution, or where the use of RPA may result in inaccurate results. For instance, RPA is unsuitable for processes that deal with unstructured data. In those cases, RPA needs to be used along with technologies such as OCR and natural language processing (NLP) that will enable RPA to read, extract and convert data into a format that is machine-readable.
Another area where RPA often fails is when it is used with processes and systems that are frequently updated. Because the majority of RPA applications do not use machine learning (ML), they are unable to respond to changes in the processes and applications that they are scripted to work with.
As such, Arendtsz told CMSWire that RPA is complex to manage at scale. “RPA tools are a collection of scripts that cannot adapt to changes. Updates or replaced applications in a process will result in a failure or incorrect data,” said Arendtsz. “Troubleshooting and error management of automations is generally a challenge, as only a few enterprise-class solutions provide robust governance capabilities.”
Businesses with many RPA processes that work in conjunction with one another are especially prone to problems with system updates. Additionally, processes that are too complex to begin with are typically not suitable for RPA, as it adds yet another layer of complexity to the process.
Unwanted Side Effects
Decreases in Employee Satisfaction
Matt Watson, co-founder of Full Scale, a tech-enabled service provider, told CMSWire that one of the side effects of RPA is that although it is able to handle the mundane tasks that employees regularly perform, it is unable to perform the more difficult tasks, potentially harming the employee experience in the process. “One potentially unanticipated outcome is that when you automate simple tasks, it leaves all the challenging work for the humans. Imagine if you work in a call center, and new automation protocols are suddenly able to resolve 90 percent of the calls that come in. Now every call that gets through to you is angry and your job just got a lot harder and more unpleasant,” said Watson.
Although RPA is able to accomplish many tedious and mundane day-to-day tasks, it may impact the satisfaction and performance of employees. “Employers need to be aware of how introducing RPA could shift the day-to-day experience of their team. They should encourage feedback during the transition, provide support, and understand that processes and guidelines that may have worked well for the staff previously could need to change,” Watson suggested.
Difficulties in Adapting to Changing Circumstances
Duri Chitayat, CTO at Safeguard Global, a global workforce management company, told CMSWire that in people-centric industries like HR, RPA has shown to be problematic. “For example, many HR tech stacks today (unfortunately) are integrated using RPA. This seemed like a good idea at first — it helped automate administrative tasks like data entry and saved workers from repeating tedious tasks,” said Chitayat. “However, organizations quickly realized that with RPA, their HR tech stacks became tightly coupled and brittle, leading to difficulty in adapting to unpredictable circumstances.”
Chitayat explained that RPAs can break or cause challenges when changing vendors or integrating new apps, resulting in having to test and update thousands of different automations, wasting time, money and resources. “RPA is designed for low-code users. The abstractions limit advantages to adopting new technology, particularly data-intensive applications.”
Final Thoughts on RPA
Robotic process automation can be useful for reducing the number of tedious, routine tasks that would otherwise have to be accomplished by employees. Because they are low- or no-code platforms, they can immediately be put into action by nonprogrammers, however, security should be taken into consideration throughout the process. Additionally, the complexity of a brand’s overall software stack may impact the effectiveness of an RPA solution.
