A major buzzword these days, a citizen developer program (CDP) or "democratization of automation" is a powerful device available in the corporate toolkit to scale up enterprise automation, particularly a Robotic Process Automation (RPA) program. 

When done right, it can bring great benefits, increasing workforce productivity and engagement, reducing risk and cost, speeding up value delivery in the enterprise, increasing ROI, and disseminating RPA adoption. 

However, CDPs will often be created without a proper strategy and implementation roadmap, leading to poor execution, frustration, and disappointing results.

In this article, we share our learnings and observations from multiple real-life experiences to help you set up a successful RPA citizen developer program, raising awareness of pitfalls and listing actionable steps.

 

Why establish a Robotic Process Automation Citizen Developer Program?

According to Gartner, a citizen developer is an employee who creates application capabilities (robotic process automation, in this case) for consumption by themselves or others, using tools that are not actively forbidden by IT or business units.

A citizen developer is a persona, not a title or targeted role. The best citizen developers tend to have structured thinking and a tendency or inclination to explore new digital technologies.

RPA Citizen Developer Programs provide the opportunity to a significant number of employees that are not in traditional IT job roles to learn an exciting, relatively easy-to-use technology and to optimize their business processes without having to rely on IT staff and their development timelines, priorities, or budgets. 

CDPs have the potential to enable quick productivity gains without the cost of adding expensive IT resources while giving business users control over their priorities and delivery timeframes. IT will be a key partner in this journey, providing guidance, oversight, and support while retaining control of the platform. 

Citizen developers may use this platform for themselves or for a department or a larger number of users while following the prescribed different governance guidelines. A CDP may also open doors to new career directions.

The low code/no-code capabilities of major RPA platforms (UiPath, Automation Anywhere, Blue Prism, Microsoft Power Automate, and others), easy-to-learn and use, intuitive platforms, and readily available online self-help are major factors that make a CDP viable for many RPA projects, particularly simple ones. 

Projects with medium to high complexity are suitable only for experienced citizen developers, while highly complex ones are better left to dedicated RPA developers.

 

First things first - establish a traditional RPA program

The proven approach for establishing a successful citizen developer program is to start with a traditional RPA program with dedicated resources. For firms starting their RPA journey, a CDP should be part of their overall RPA strategy and roadmap, to be added after establishing a solid foundation. 

For RPA programs that are well underway, a CDP can be added, and the overall strategy adjusted accordingly, leveraging the lessons learned and adjusting policies and procedures of the traditional RPA program. 

During the early phase of your program, you may start planning for the impact of CDP on the following:

• Bot Ids – Identity and Access management Ids to be created exclusively for bots
• Policies and Procedures approved by Compliance, Audit, Risk, and Legal
• Bot monitoring procedures
• Training – Training programs created for citizen developers can be adapted from the training provided to RPA developers and may also be tailored for different citizen developer profiles – developers, power developers, bot managers, etc.

 

Start Small and Scale up

The RPA CDP should be rolled out similarly to the traditional RPA program: start small, select one business team or department, and scale it up afterward. Some key elements are:

• Establish a dedicated Citizen Development Program team within the Center of Excellence
• Identify potential processes to be automated, categorizing and prioritizing the opportunities by complexity and expected benefits
• Establish a “Community of Practice” for a wide audience
• Identify and train potential citizen developers
• After training, the citizen developers will start their RPA journey under the guidance of RPA COE
• Identify those citizen developers who could influence their teams as part of the CDP, nurture them, and encourage full participation in a community of practice.

Establish the Citizen Developer Program

A dedicated team within the RPA COE, focusing on specific citizen developer needs while providing oversight for policy compliance, vendor relationship management, RPA software upgrades, infrastructure, platform upgrades, and overall support and enablement, is critical for a successful citizen developer program.

There are multiple new roles required for CDP. These roles would ideally be performed by fully dedicated resources, but some of them may be fulfilled by core RPA COE team members wearing multiple hats:

• An RPA COE Lead may take on the additional responsibility of RPA CDP Program owner, an overall owner of the CDP program, with overall responsibility for tracking and reporting the benefits of CDP to senior management by managing relationships with different business units participating in the CDP
• RPA Business Analysts (BAs), as process intake/discovery owners, will work with the business units initially to help them identify and prioritize opportunities (if needed), create appropriate documentation, verify test cases and test data, etc.
• Project Manager manages the program, monitoring the status and effectiveness of each project and of the program as a whole
• RPA CDP Platform Manager manages and permissions in the RPA platform, code migration between environments (Dev to QA to UAT to Prod), application installation/set up, monitors bot activities, and performs other technical tasks
• Development and support engineers help to:
  • Create and publish standards (naming conventions, coding standards, versioning, etc.)
  • Oversee the CDP training program and mentor new developers as needed
  • Review and approve process designs (requesting design revisions, as needed)
  • Review code prior to moving to UAT and Production (utilize code review tools if available), prepare for platform migration, and promote code to production, as needed
• Communications Manager manages the Community of Practice to communicate with citizen developers, businesses, IT, and across the organization; maintains an internal site – using SharePoint pages, the confluence of similar tools -; publishes, monitors, and responds to internal social media / Yammer.

 

Implement Policy or Guidelines for Selecting Citizen Development Projects

Establishing clear guidelines to identify which processes can be developed by citizen developers and which ones should be routed to a traditional RPA development team is critical for a successful CDP

• For simple processes, governance, development standards, and procedures should be light, reflecting that simplicity. This will enable a significant number of citizen developers to develop, deploy and maintain their own processes
• For more complex processes, more governance is required. It is better to handle those automations through a traditional RPA development team. As the citizen developer program matures, dedicated resources (internal or external) may be identified to work in a hybrid model, following CDP procedures.

 

Establish Tollgates

RPA CDP should establish an appropriate level of governance and tollgates:

• For simple process automation, a use case coupled with a simple process description or flow may suffice. The documentation should be reviewed and approved prior to going live
• Prior to deployment to production, the code should be reviewed to ensure that it works as designed, requirements are met, and good/approved standards are followed – naming convention, coding standards, etc. Changes should be implemented as needed, following change management procedures
• Testing requirements should be established – required test data should be available for testing. Test results should be stored for audit purposes
• Formal approval process needs to be established for each stage.

 

Guidelines and rules should be established for:

• Completion of UAT
• Completion of supervised production
• Change management and maintenance
• Code fixes

 

Monitor Bot Activities

Bot activities should be monitored to ensure that bot licenses are being fully utilized, benefits are realized, and issues are being addressed

• Bot licenses that are not utilized after deployment should be investigated for redeployment if they are no longer needed
• Unattended bots that are used within narrow time slots (and low utilization rate) may be able to perform other tasks during existing idle time slots
• Bots that are failing often may need to be redesigned, redeveloped, or retired
• Bots may be performing unauthorized actions. Bots that are not following approved procedures may need to be redesigned or prevented from running until violations are resolved.

 

Validate the Benefits and Value Delivered

Early in the program, it should be defined how the benefits will be measured for the RPA CDP. Multiple metrics should be developed and may include:

• Number of bots deployed, or number of processes automated per month or quarter
• Number (or percentage) of transactions processed through automation – per month/quarter/year
• Estimated number of hours saved per month/year
• Other quantifiable and unquantifiable benefits like revenue lift, SLA improvement, accuracy improvements, rework avoidance, overtime avoidance, regulatory compliance, and employee morale should be considered and incorporated into the overall benefits
• ROI – cumulative and monthly/quarterly estimates

 

Establish a Community of Practice

Creating a community of practice, using a forum-like environment curated by the COE, for internal RPA CD teams to share experiences, suggest solutions, and receive COE communications, is crucial for the long-term success of the citizen developer program. 

Yammer and similar tools will keep the buzz going, enabling the community to share success stories. This dynamic will also enable the RPA CDP COE to gauge the level of satisfaction, address potential pain points, and causes for frustrations – and address them in a timely manner to ensure the success of the CDP.

 

Conclusion

RPA citizen developer programs have a greater chance of success as an extension to traditional RPA programs, with their own governance and committed resources.

Policies and procedures, standards, tollgates, roles and responsibilities, and monitoring tools should be established early in the process, led by a Center of Excellence comprised of experienced RPA practitioners, with internal or external resources, as needed.

 

This article was contributed by our expert Hiren Kotak

 

Frequently Asked Questions Answered by Hiren Kotak

Q1. What is an RPA program?

Fundamentally, an RPA program is a set of activities to implement Robotic Process Automation (RPA) software and utilize it to automate a high-volume, repetitive, tedious manual task. This program is usually run by an RPA Center of Excellence or Operations Excellence department, with a close partnership of IT groups such as Infrastructure, InfoSec, and Access Management. 

In order to find the right type of opportunities for RPA and prioritize them, a group of process analysts or business analysts perform a process discovery, feasibility analysis, and cost-benefit analysis while working with Process Owners, Performers, or Subject Matter Experts (SMEs). RPA Developers utilize the process documentation to configure the software robot (“bot”).

Once the SMEs fully test the bots, they are implemented in a production environment and closely monitored. Most bots don’t handle all possible scenarios, so some of the complex or less-frequent cases are still handled by staff members. The Bot controller monitors the bot’s performance and triages any issues that may come up. 

 

Q2. What are the types of automation in RPA?

RPA is generally categorized into two types: 

a.    Attended (also known as Desktop Automation, triggered by an individual)  
b.    Unattended (scheduled or triggered by an event)

Attended automation is useful for individual task-level automation on one’s desktop or laptop. Usually, this automation serves as a ‘digital assistant’ to complete some of the repetitive, tedious tasks, which could be a part of the overall process that the user (bot owner) would perform. 

Unattended automation usually serves an entire group or department, as it would process transactions without any manual interventions. They are scheduled on a particular day(s) or time(s), or they can be triggered by events such as ‘file gets created in a folder,’ or ‘an email arrives in an inbox.’ 

These automations usually run as a batch process and prepare a report of successful and unsuccessful transactions for the process owner or supervisor to take appropriate actions. 

 

Q3. How does Robotic Process Automation (RPA) differ from Intelligent Automation (IA)?

RPA, in its purest form, is the automation of rule-based tasks with well-structured input, well-defined business process rules, and associated actions. Many RPA tool providers have now introduced cognitive components in their platforms that can handle semi-structured input and probably work with somewhat loosely defined rules. 

Intelligent Automation is the automation of tasks that involves the cognitive capabilities of the human brain, i.e., processing of unstructured data (e.g., free form text), making judgment calls, and so on. It involves actions based on probabilistic analysis and some level of ‘predictions.’

In order to achieve end-to-end process automation, we commonly use both types of automations, and they are sometimes collectively referred to as hyper-automation. 

 

Q4. How does Robotic Process Automation RPA interact with applications?

The beauty of RPA is that it’s not invasive, which means it does not involve traditional software development with a new user interface, batch processes, coding, database, and so on. Traditional software development forces users to adapt to the prescribed new way of performing tasks. 

RPA keeps the original process as is (with minor changes to digitize input or make it RPA-friendly), does not change the underlying platforms or applications, and hence requires practically no new training for the staff.

RPA, again in its purest form, uses a user interface of applications, just a way a human user would use it, i.e., bots or digital workers login on to an application just like another human being (maybe a superhuman being!), and performs the same tasks that human workers performed. It mimics or simulates all human actions to get the desired output. 

In recent years, some tool providers have included API-based actions in their tools, making it a more robust low-code, no-code automation tool with in-built connectors and pre-defined actions.