How to Build a Successful RPA Program
The goal of companies that have robotic process automation (RPA) programs is to become self-sustaining. When starting out in RPA, an organization may need the help of an expert consultant, but it ultimately aspires to succeed on its own.
Business strategy for RPA has been talked about for almost a decade, but few companies are taking advantage of it. Why?
Today, companies know what software robots are. However, RPA needs to be judged by how much it has penetrated the market and how much it has penetrated the companies that are using RPA. Do companies have an automation mindset?
RPA has done a good job penetrating the market. Half of the 20,000 largest companies are using automations, but the RPA programs at these companies are not very effective. Most organizations are not seeing the benefits they should be seeing or even the benefits they expected.
Here’s a look at the mistakes companies are making and critical factors for building a successful RPA program.
A Common Mistake
Some companies may have profitable programs, but they don’t see the 3x, 4x, or 5x pay-off seen by the customers with the most successful programs. The core reason for lower profitability is that a lot of organizations approach RPA as a series of projects. They ask, “Which bot am I going to build now and which am I going to build next?”
This approach is a big mistake. The company may start out by getting a process assessment and coming up with a process catalog, which is an ordered list of opportunities for automation. The first thing the company does is sort these opportunities in descending order of benefit. They point to the first one and ask for it. Each project, even if it brings good returns, is not going to move the needle for your company.
Characteristics of a Successful RPA Program
What moves the needle for your business is having an RPA program within your company that is:
a) Robust for identifying opportunities for automation within your company and has an automation mindset
b) Able to build the bots that implement these opportunities
c) Able to run these bots at scale
To achieve appreciable results with RPA, you need a lot of opportunities that produce ROI. Reaching this point requires treating RPA as a practice.
If your company is treating RPA as a practice, you are going to take the process catalog and sort it by how difficult the RPA processes are to implement. You should build the easiest automation first. To scale, you need to build your automation development muscle. It’s like lifting weights. If you want to reach a weight-lifting goal, you need to start with what you are capable of.
Critical Success Factors for Building an RPA Practice
To successfully build an RPA practice, your company needs to retain the brain. Bots need to be built in house and you must have the automation mindset in-house.
The number one critical success factor is data discipline. For the entire lifecycle of the automation, you need to know the benefit and the cost. A consultant does this on the front end, and that’s how to prioritize automations, but organizations forget this as they start to implement automations.
Companies need to make a process definition, and by the end of the document, you might decide you don’t like the ROI. The benefit may erode down the lifecycle. You know the cost upfront but not the benefit. Your company needs to understand ROI at the bot level to get to ROI at the program level. You should be able to speak to the benefit of the program off the top of your head to justify it.
Other critical success factors include:
Building a healthy automation pipeline through automation assessments that generate automation opportunities
Cultivating an automation mindset through training that reduces fear and uncertainty, giving employees a chance to nominate opportunities, and cultivating these opportunities
Implementing change management with the understanding that your company won’t see benefits right away
Structuring to scale because establishing a proof of concept (POC) with one bot doesn’t get results. Build the first bot so that, if you add more, they can still run, and you can scale to 50 to 100 automations.
Resourcing, the one thing in common for all sizes and types of companies. The right talent must be pointed at challenges when you need it. Certified developers are rare. Onboarding takes time, so you need to plan for this. Resource needs change, and getting a mix of the right talent is difficult.
What Success Looks Like
The timeline to RPA program maturity should signal to you whether your program is working. Your company needs visibility into 50 bots by the time you’ve been working at an RPA program for 8 months. If not, you are missing out on opportunities to free up human capital.
Customers can scale from 0 to 150 automations by the time they’ve been working at a program for 12 months. Your practice should feel like driving a high-performing car.
Analysts are talking about RPA in new ways, labeling it as hyper or intelligent automation. Analysts are pressured into coming up with new ways to package something, but the same RPA is at the core: software that mimics human action. RPA by itself is limited in applicability. Companies use it to automate manual processes that software could be doing better to take these processes off the plate for humans. Repackaging RPA can expand the real estate of what a robot worker can do.
Companies need to expand their use of RPA. As the world changes and new technologies come online, they need to be integrated in an effective way. RPAs are a great way to pilot a new technology before you know it will work in your legacy system.
As knowledge workers are freed up by RPA, staff can be retrained. Probably the most important skill in the next 15 to 20 years is going to be computer programming. There is a value to knowing how things are done. Computer programming will become similar to knowing how to read in the 19th century. It will give people access to a whole new economy that is emerging now.
RPA is part of the way we democratize this knowledge. There are challenges to robots taking over human work, but ultimately, it’s a good thing. RPA frees human capital to do other things, positively affecting prices and what people can do with their lives.
How to Get Going
Even if your company doesn’t do anything with RPA, it’s never too late to get started. Get informed about what the industry does and develop a clear vision of what you can accomplish.
Look into data discipline. You should know how big your pipeline is and, on the development side, how much throughput you have. If you have bots in place, understand how much benefit is available and hold yourself accountable.
It helps to talk to partners that have been around for a while. It doesn’t cost a lot to have a conversation.