Amidst the rapidly changing business dynamics and customer expectations, most organizations have been looking for ways to improve business efficiency and gain a competitive edge. Process automation is an ideal approach for organizations to make their processes more efficient. Automation can reduce manual involvement in both internal/backend as well as customer-facing processes, making them faster, consistent, and more accurate.
The benefits of automation make it a compelling solution, however, not every process may necessarily be a contender for automation. To begin with, we look at the steps an organization must take before initiating an automation project. Overall, the Automation Roadmap covers the steps identified in the image below:
Pre-Automation Process Mapping
To make improvements to business processes through automation, organizations first need to map all components and elements involved in a way that exposes all inefficiencies.
Modern business processes are an ecosystem where various components, such as people, systems, workflows, data, information, and business rules, constantly interact with each other. (Refer to our article on Integrated Business Process Engineering discusses for more details.) An automation project may potentially impact this entire ecosystem, and so before embarking on your automation journey, it is important to model the business process ecosystem and the interaction between all the elements.
While there are several ways to create process maps and identify inefficiencies, which in turn become opportunities for automation, they typically fall into two approaches: bottom up (or Business Process Identification) and top down (analysis of organization’s KPIs).
Business Process Identification
This is a bottom up approach, which includes two distinct processes: Automated Process Mining and Automated Process Discovery.
Automated Process Mining is a process by which we run various specialized programs that crawl through all dataflows and event logs in an organization and build process maps. This automated analysis captures all system interactions and is "noise-proof”, but it does not capture user-only based tasks.
Automated Process Discovery, on the other hand, is a process which captures user keystrokes on all workstations and analyses them for process identification. While this method captures all user-based tasks, it may not capture the entirety of systems-only tasks. It is also susceptible to noise from non-work-related keystrokes.
Analysis of Organization’s KPIs
This analysis allows us to look at the materiality of inefficiencies by evaluating the revenue and profitability KPIs on one hand, and expense side and operational excellence KPIs on the other. Revenue and profitability KPIs demonstrate process effectiveness, and expense and operational excellence KPIs show process efficiency. Organizations typically need a well-developed KPI framework to conduct a root-cause analysis and attribute, for instance, an increase in expenses to a specific KPI inefficiency or to ascertain whether an observed inefficiency impacts customer experience or profitability.
In order to understand which processes are the best candidates for automation, we need to look at all the above three approaches – process mining, process discovery, and KPI analysis. Once we understand both system and user-based tasks, the process inefficiencies, and the materiality of those inefficiencies, we are ready for Robotic Process Automation.
What is Robotic Process Automation?
Robotic Process Automation (RPA) is a technology that replicates human interactions with systems (operational, analytical systems, modular components, applications, etc.) and websites by:
- Extracting, adjusting, and recording data through data and web extraction frameworks or screen-scraping methods
- Execute pre-configured business rules in operational contexts
- Applying data quality checks and adjustments (eg. Data standardization) based on pre-configured rules
- Invoking other systems and services, and exchanging data with them, based on pre-configured business rules
- For instance, extracting data from systems and pulling it into tools like Excel and preparing the content for downstream analytics
Essentially, the automation engine can be thought of as an orchestration tool that can coordinate the interactions with multiple systems to execute a series of tasks as constructed within the automation workflow.
Will RPA replace human workers?
This is a question that has sparked debates across the world. However, even though RPA is essentially structured to replicate human activity, it is not equivalent to human workers. Standalone RPA technology cannot:
- Learn: Unlike humans, the technology cannot robustly adapt to new events or expand/improve on its existing decisioning capabilities
- Reason: It cannot objectively assess validity, reasonability, and completeness of data outside of static business rules. RPA technology makes decisions based on the logic you provide, but some intelligence can be added through AI-driven automation
- Rationalize: It is unable to make rational assessments and judgments about events and findings outside of static business rules
- Evolve: It does not get more proficient over time and cannot dynamically adjust or optimize the business process based on experience
In fact, a recent study by Statistics Canada on the Effect of Robots on Firm Performance and Employment found that organizations that had invested in robots and automation since the late 90s have also expanded their workforce. The study noted that firms that had automated some processes had 15% larger workforces than their contemporaries in the same industry. Organizations typically automate their tasks to improve product and service quality or process efficiencies, rather than to reduce labour costs, and this helps in further expansion of the business as well as in volume growth within the automated processes. However, some low-skill jobs may give way to more strategic roles and other jobs could change with the adoption of automation.
Enablers for Successful Enterprise RPA
There are a number of enablers for successful RPA deployment within an enterprise. For one, the corporate strategy should be aligned with RPA and should be aimed at rationalization and simplification. It should also account for and handle the investments needed to migrate to automated processes, maintain them, and allow for reasonable experimentation.
The enterprise data needs to be well-understood, consistent, correct, and stable. This usually requires having in place a data governance framework, suitable processes for metadata management, and a data quality engine. This data also needs to be accessible by the RPA workflow, typically through a service account or other types of elevated privileges.
To minimize automated process failures, the technologies being interacted with also need to be understood and documented at the application level. The limitations need to be documented and access needs to be configured for RPA. Finally, the business process needs to be rationalized, optimized and well-understood, as the RPA tool will essentially replicate the current processes. Process efficiencies should focus on ensuring tasks are made as sequential, stable, and minimalistic as possible to reduce potential points of failure in the automated workflow.
Assessing Candidates for RPA
When an organization assesses the candidacy of its processes for RPA, there are two major things to consider: Viability and Priority.
Assessing the viability of a particular task for RPA would entail determining anticipated return on investment (ROI). This can be calculated by determining the time that will be saved by automation, user errors that could be mitigated, the effort require to gather input data, the scope and variability of the process, and whether the input data and process would support automation.
Prioritization is usually a factor of complexity and value of the tasks. For instance, low complexity tasks that deliver high value will naturally be prime contenders for automation. Low complexity but low value tasks usually offer a learning opportunity and can be completed in the shorter-term. In the longer-term, organizations can focus on higher value but high complexity processes, while high complexity but low value tasks would be the lowest priority and they might either be looked at as an exploratory challenge or may never be automated.
Solving Business Problems with RPA
Business applications of RPA vary within different organizations, but these tools usually focus on one of two overarching themes:
Repetitive Workflow Automation: This encompasses tasks that follow a repetitive workflow, such as submitting forms/requests or capturing and summarizing information from different systems (eg: extracting static information from standardized documents like tax forms, etc.).
Communications & System Integration: This goes beyond the application layer and enables the tool to trigger and send notifications/emails or update or create reports or forms.
One of the major benefits of RPA tools is a reduction in resource utilization and time requirements for the tasks being automated. It is also an opportunity to improve process quality and consistency and simplify and standardize processes. It is useful in enhancing compliance and traceability, as RPA tools follow a rigid, rule-based structure and reduce the risk of human errors. Moreover, as part of the RPA flow, you can choose to document the steps and produce a process log. Through exception handling, the tool can also send out email notifications in the case of process anomalies, allowing for both quality improvements and time saving.
Going Beyond RPA with Intelligent Automation
While RPA, by itself, has the potential to transform business processes for many organizations, it is by no means the pinnacle of automation. RPA processes are typically not 100% automated unless they are extremely simple, as there are often high complexity events that cannot be accounted for in this type of automation. Some business needs may call for Intelligent automation, which is a customized automation solution that combines the right RPA and AI capabilities to drive automation beyond simplistic, rule-based logic. Artificial intelligence can augment the automation flow and enable semi-autonomous decision-making and unstructured data processing. Intelligent automation is already being used by organizations across industries for pattern recognition, determining probabilities of event occurrence, anomaly detection, and AI-driven chatbots, among others.
Not all workflows, however, require added intelligence. Some considerations to be made for whether or not to incorporate AI to automated workflows include workflow complexity, variability, data quality, business criticality, governance and compliance requirements, and customer impact. As the organization undertakes comprehensive process mapping, the process inefficiencies and some of these considerations will be made clear. For simpler automation requirements, RPA is a good place to start as you work to improve business processes and increase efficiencies in line with your organizational strategy.
Exploring Robotic Process Automation or Intelligent Process Automation to enhance your organization's process efficiency? Our experts can help you identify ideal candidates for automation and implement RPA/IPA. Connect with us to learn more.
Book a Free Consultation
We will contact you as soon as possible.