Robotic process automation: getting your implementation plan right first time

Robotic process automation (RPA) in its simplest form enables rule-based processes to be automated. As a result, employees can refocus their time on higher-value tasks. But, while many organizations are keen to add RPA to their digital transformation strategy, some are falling at the first hurdle. While RPA might seem like the right solution for your business, it can take time to realize value if not implemented carefully.

So can you get RPA right first time? The short answer is yes, but you need a plan. Here are six steps to help your organization get started with RPA.

An essential part of any project is understanding why there is a need for the project in the first place. Once you agree on your objectives, everyone involved becomes united by a common goal. In this scenario, you need to discuss what you hope to achieve with RPA. Are you looking to reduce costs? Do you want to speed up processes? Are you hoping to reduce human intervention? Or is it a combination of these things and more?

Whatever your objectives and measures of success are, they need to be achievable. For example, a leading brewer decided to automate part of its order-management process to improve its critical "on-time, in-full" supply-chain metric. Without this clarity, implementations can be considered failures purely because the objectives were unclear from the start.

This might seem obvious, but it's incredibly important to define the roles and responsibilities of every individual involved in the RPA project. Not just once the project kicks off, but before, during, and after. Everyone has a day job, so they need to fully understand what they are committing to and the effort required. And, don't forget, you're likely to need some of your experienced IT professionals aligned on the project, too. Plus, your HR, learning and development, change, and digital management teams should also be involved from day one.

Who are your key stakeholders? And do they understand what is and isn't achievable? For example, although you could theoretically have a 1:3 man-bot ratio, often only 1:1.2 or 1:1.3 is possible. This is just one example of why it's crucial to ensure your objectives and measures of success are clearly established at the start.

It's also important that you agree on the levels of automation you expect to achieve across different processes or where you won't be automating at all. A leading healthcare business ran a successful automation program by creating an automation committee, comprised of business leaders, who met on a weekly basis to discuss the project. Only when you have all the relevant humans aligned are you are ready for the robots.

There are so many options when it comes to RPA, but if you select the wrong process, there's a high chance you may not meet your objectives. Therefore, the assessment and design stages are critical. Firstly, review each process objectively for automation potential. Then, examine the ROI of each process and run a feasibility study. Also, consider other ongoing or future projects that may impact the process and fix any existing breakages as soon as possible.

Lastly, sort out license requirements and approvals before implementation. I would agree with the rule of thumb followed by one of our automation clients in pharmaceuticals: at least a third of the bot investment should happen in the assessment phase.

While you might want to jump straight in with RPA, remember that slow and steady wins the race, especially during your first few implementations. Of course, you should establish an end-to-end intelligent automation roadmap, but it's good practice to start with a pilot. This will help you spot any small issues before they become big problems. Then, when you're ready, you can scale up in phases.

For example, one of our retail clients started with 10 bots in HR in the first year, which increased to 20 bots in the third year. Today, they have a model for over 50 bots annually that include artificial intelligence and machine learning components.

By step six, your organization is enjoying the benefits of RPA. But you should never settle – there's always something to be improved. Keep an eye on best practices, including reusing and standardizing code. You should also leverage heat maps of existing bots to increase the man-bot ratio where possible. It might be tempting to sit back and marvel at what you created, but RPA is an opportunity for continuous improvement.

By following these steps, you can set up your RPA implementation for success. When it's time to advance into intelligent automation – by adding advanced technologies like AI, machine learning, natural language processing, and analytics to your program – you'll be ready.