This past year has been challenging for businesses of all sizes, and many companies turned to technology solutions for help—specifically, Robotic Process Automation (RPA).
It makes sense. The COVID-19 pandemic has drastically changed the business environment we operate in. Companies have needed to quickly and meaningfully adapt their operations. They’ve had to accommodate disruptions in circumstance and changes in what customers are asking for. They’ve also needed to find ways to make more efficient use of their technological tools and systems, better orchestrate processes, and streamline lower-value busy work so their people could focus on delivering value. For the most part, that meant accelerating digital transformation projects from planning, testing, or pilot to full production deployments. And in many instances, RPA was the technology solution companies turned to.
What is Robotic Process Automation?
Robotic Process Automation was developed over a decade ago as a method of automating tasks and mimicking how human workers (often referred to as knowledge workers) interact with desktop applications. The overall goal was to streamline how work is done so that people could focus on higher value tasks. The automation solutions are built using software scripts that are referred to as software bots, or simply “bots” for short.
Like all things, RPA has evolved in sophistication over the years. Today it includes a bot-building, graphical drag-and-drop interface that requires less coding than before. From these humble beginnings, innovations in RPA bot building have advanced to include the ability to build bots using new screen recorders that anchor on underlying web objects or technology such as computer vision for object detection.
The companies that are successful with extensive RPA deployments typically focus on tasks of a low-to-medium complexity level. These deployments often entail putting sophisticated and complex code snippets from a traditional programming language such as Python in their bots. By having RPA features and functionality focused on task automation requirements, RPA has the ability to create bots that can address most simple automation needs.
One straightforward RPA use case is simple “swivel chair automation”—which, as the name implies, mimics the work of a person “swiveling” one’s chair working between applications to move or process data between them. Consider the finance and accounting (F&A) 3-way match invoice processing that companies perform prior to invoice payment. Various validation rules or checks need to pass prior to payment occurring.
- Is this a duplicate invoice that is already paid?
- Were the goods or services received and complete?
- Is payment approved by the requester?
An RPA bot can be used to automate the task of validating these rules and passing the checks. Such tasks are within the sweet spot for RPA bots. They are well-defined, deterministic, and are not subject to the need of constant change or modification.
When is RPA not the right fit?
While RPA is great for streamlining individual manual and menial tasks, to do something more complex, such as improve a process end-to-end, or optimize your operations more holistically, you need a solution that’s more comprehensive and accessible to everybody in the enterprise.
Simply put, RPA was built to facilitate task-level automation. It is exceedingly difficult to use RPA bots to facilitate entire processes end-to-end. The current method entails trying to run a series of bots back to back.
It is technically possible for a developer to string together bots and using various event triggers, schedules, and bot master/child relationships to create a pseudo workflow that mimics a business process, but there are issues:
- Only highly technical bot developers who understand the tool and the programming principles can maintain the bots.
- The business logic is embedded into the bots, making it impossible for business users to maintain.
- End-to-end business process maps still must be maintained external to the RPA system.
These restrictions encumber the agility of the business, and undermine your ability to quickly refine and update how people work. Business owners often need to deploy manual workarounds while they wait for the highly trained developers to update the bots for any process changes.
Take for example the employee-onboarding process. This process is composed of various tasks that need to happen in a specific sequence across multiple days. These tasks may require various approvals and actions along the way, such as the need to do things like:
- Initiate a background check immediately after offer acceptance
- Create IT requests for computer equipment ordering, assignment, and provisioning
- Create facility requests for space assignment, badge and physical access controls
- Request, approvals, and provisioning of job specific applications access (Salesforce, Jira, SAP, etc.)
- Assignment of mandatory trainings (Culture, Safety and Security, Travel and Expense, etc)
Designing and building this HR onboarding process would be inherently challenging using RPA. Plus, as the HR onboarding process evolves to include additional items such as buddy assignment, welcome lunches, COVID protocols, etc., maintaining it would be time-consuming, as all of those changes need to be made by the RPA developers.
The complexity of the bots, coding constructs, and inbuilt error handling is often too much for citizen developers to work with. To address this, most companies will enlist a group of developers to manage this ongoing maintenance. Unfortunately, this creates a big bottleneck, which is costly, especially to mission-critical processes that require fast updates to maintain business agility and a competitive advantage.
Luckily, there is an easier way.
This is just one piece of a larger puzzle.
RPA should not be used as a silver bullet. Rather, it should be used to augment humans and workflows and used as a tool in a larger toolbox or one piece of a larger operational puzzle.
Problems related to inefficiency, adaptability, and operational complexity are inherently cross-functional and complicated in nature, and thus ill-suited for RPA. A proper solution must be holistic and multi-faceted. The solution needs to be more powerful and comprehensive, extending beyond task automation. Something more like process orchestration.
Tonkean was built with this in mind. It is a comprehensive and holistic platform that allows orchestration of your operations’ various moving parts—operations unique mix of people, systems, and tools—such that you can improve each symbiotically, in a manner that truly enables the business to break down the bottlenecks that interfere with efficient work. Tonkean also enables companies to use powerful tools such as RPA strategically. Inside of Tonkean, IT teams can provide the business building blocks and technical capabilities—from RPA bots to API connectors—which the business can utilize in the no-code workflow solutions built inside Tonkean Solutions Studio. This gives the business greater agility, by allowing operations teams to rapidly modify processes as needed —without ever having to inconvenience IT. Tonkean is an operating system for business operations enabling the business to run smoothly and efficiently.
There is a saying about the importance of using “the right tool for the job.” RPA is well-suited for building out task automation, but its utility begins to diminish when you ask it to do things like support end-to-end process improvement or process orchestration.
For that kind of work, you’ll need a platform designed to help you make better use of all the tools in your toolbox.