Robotic Process Automation (RPA) has rapidly become one of the hottest trends in enterprise technology, with the potential to significantly drive down enterprise costs while improving productivity, quality, and compliance.
However, achieving the intended results from RPA goes far beyond building and deploying a bot. And with the potential to deploy hundreds of bots at once, a haphazard implementation can considerably diminish RPA’s potential. However, with a compelling rationale for change, a well-defined business case, and applying implementation and governance best practices, business and IT leaders will be well positioned to drive successful RPA outcomes.
There are three main RPA types. The first is characterized by simple screen scraping, workflow, and other similar tools that automate discrete tasks – for example, copying and pasting data from multiple sources and aggregating them into a central repository. These automations are relatively unsophisticated and developed for narrow use.
Next are more robust, general-purpose tools that drive the majority of current implementations delivered by automation firms, such as Blue Prism, UiPath and Automation Anywhere. Implemented through bots, these automations typically mimic the manual path taken by a human to complete a particular process. Usually applied to stable, high volume, and rules-based processes, these tools can interact with multiple enterprise systems and can be deployed relatively quickly.
Lastly, there are tools that utilize higher-level artificial intelligence and machine learning. While still relatively nascent, these technologies are being adopted to tackle cognitive heavy processes, such as those that require unstructured data processing. Examples include IBM’s Watson, Genpact’s RAGE, Antworks’ Antstein, and IpSoft’s Amelia.
Organizations considering the deployment of an automation program should consider the following guidelines:
Start small and “test the water”
If your organization is just starting its RPA journey, consider a proof of concept to rapidly confirm benefits and identify potential issues before the full implementation. Identify two or three processes that are simple, have well-defined rules, use structured data, and are stable.
This step provides valuable insights that help fine tune the broader implementation. This test run provides reasonable confirmation that the advertised benefits are applicable to your organization; identifies constraints that hinder implementation including applications access and information security restrictions; and socializes RPA’s benefits to key stakeholders to create automation credibility and engender wider support.
Formalize next steps and get executive leadership support
Once the preliminary support is obtained, formalize a roadmap to gain senior leadership’s support for an enterprise wide implementation. Following are seven key steps to establishing organizational readiness for RPA and securing critical executive backing for an RPA program.
Dimension the “size of the prize.” Perform a comprehensive process sweep to determine the total opportunity size and the implementation level of effort (including processes that should not be automated). These opportunities should be listed in order of priority.
Develop your automation strategy. Articulate the vision and rationale of the RPA transformation and the outcomes you plan to achieve. Consider functional appetite, roll out strategy (center of excellence vs. functionally led), and the tool selection approach in developing this strategy.
Develop a comprehensive business case. Ensure the total benefits justify the needed investments. Quantify the benefits including savings from resource rationalization, decreased rework, reduced overtime, and increased compliance. Additionally, quantify key cost drivers including provisioning of RPA licenses and other IT resources, and any external services required.
Partner with IT as early as possible. Elicit the IT team’s partnership upfront to ensure tools are properly integrated; hardware, software, and labor support are provisioned; and applications access is provided.
Create a future governance and operating model. Define the right governance model to manage work performed by both human and digital workforces (e.g., what action will be taken if the bot stops?). Consider the governance mandate and the operating model that yields the best outcome (e.g. will the new RPA Center of Excellence be centralized, federated or a “hub and spoke”?). Most importantly, identify the resources and skills needed to support this governance structure.
Define the support model. Assess whether the organization has the requisite skills and capacity to implement RPA. If external help is required, determine the right sourcing strategy to acquire the support. Finally, put a plan and budget in place to build the skills internally over time.
Obtain leadership support. Make the case for change based on information gleaned from the previous steps. Ensure benefits and investments are transparent and justified to better attain executive support with the broader implementation.
Implement and industrialize
Once executive support is obtained, it is time to initiate the full RPA program across the enterprise. Start by defining the project charter, establishing the team, and developing the detailed project plan. Leaders should identify the program lead, IT and process SMEs, and other critical participants, and specify activities, dependencies, and resource requirements for each phase compliant with the high level roadmap.
The next step is to provision the IT requirements including procurement of the RPA licenses, provisioning servers, granting application access, and providing system sandboxes.Once this is complete, actively manage and communicate the shift to RPA with impacted stakeholders because automation could significantly affect how work is performed and managed. This includes proactively defining redeployment strategies as resources are rationalized.
To complete the adoption process, leadership should review and refine their automation roadmap as the deployment evolves.This includes establishing mechanisms to ensure benefits are being realized and periodically refining the roadmap based on changes in business imperatives and the technology landscape.
When it comes to beginning an RPA implementation, starting small and then expanding, while following the guidelines above, will help your team lay the groundwork for a successful RPA project that yields intended business outcomes.
Jean Paul Baritugo is a senior associate at Pace Harmon, a business transformation and outsourcing advisory firm providing guidance on complex transactions, process and operational optimization, and provider governance.