By David Brain
During the 2017 SIG Summit in Amelia Island, Florida, Symphony COO David Brain took aseries of questions from the live audience right before setting out on automating a process liveat the event. The topics ranged from business case inquiries to industry-wide observations. Have a look at how David answered the most pressing questions.
1. You suggest errors are usually a result of human errors, how can we confidently say that there are no quality issues in RPA when it is programmed by humans?
The way that the software works is that it will follow the rules exactly how you tell it to. This doesn’t necessarily mean there won’t be any errors. In the early days of RPA one of the first clients involved was a telecommunications company, which programmed a robot to send out iPhones to customers who ordered them. In the testing phase, they put a loop in order to perform some tests. When it went live, they forgot to take out the loop, so it sent everybody about 3 or 4 iPhones. That is one example where there can be issues and it can fail at scale, so you have to be very careful when testing. But there is no risk that it will just forget a step at one stage.
2. What opportunity areas have many clients gone after as low-hanging fruit and where do you see the next important ones to look out for?
RPA is not a process-specific solution, so it is not the answer to ‘how to answer invoices’ for example. What it is great at is automating the gaps in processes and systems, where you have people plugging those gaps at the moment. Where we look for opportunities are processes where there are large amounts of people performing rules-based work (service delivery centers are a good opportunity) and where that input data is structured. So, call-centers are less suited for large scale RPA and are better for the more agent-assisted solutions where you have robots on the agent’s desktop. If you are looking to have more of an impact in automating a process, look to where people are handling data at the large scale. HR is a very good area because it is typically under-invested in. Examples consist of onboarding, vetting employees, and payroll finances.
3. When you think about automation from a contact-center perspective, what are some of the challenges and pitfalls to avoid?
The largest challenge is around the systems themselves. It is important to understand how the systems work with the automation products. There are many automation tools, each with particular strengths, and some that more catered to the contact-center market. Look to see how it interfaces with the application, especially the quality of integration. If you have to go in with a looser integration, where you have to look at screen elements and rely on screen scraping technologies, then you can end up with more issues when you roll it out across a large workforce, where everybody's computers and configurations might be slightly different.
The other part is to make sure you understand the processes. We always rely on our clients to be the process experts because you know how the processes need to run. When the question is asked whether to re-engineer versus automating it, they are not mutually exclusive. We strongly recommend that if you have the time and the business case is there, to optimize the process first and then automate it. But you should also have a degree of pragmatism. If you have a process that’s good enough and can lead to large savings immediately, it can be worth it to implement the automation and refine it over time.
4. How is RPA going to impact the BPO offshore market?
All of the large BPO providers have made bold statements about what RPA will do for their businesses. For instance, some declare that it will automate 50% of the FTEs performing the processes today. These are the companies that have stated targets and business plans to increase their revenues and are trying to move up the value chain. This is where we see a real gap in the market, which we are addressing by starting to offer services around Robotic BPO (R-BPO). If you have a process that is well-suited for automation, we can provide that as a service and handle the exceptions.
5. What strategy would you recommend companies to use for RPA? Is there a situation where you build it in-house and use your own technologies and team to set it up? Or should you go outside to a specialist and have them do it, so you can avoid internal overhead and cost in the long term.
We are very open with our clients and support them with whatever strategy they decide to do. So if you want to build a team, we will support you. We run training and can help clients build their own Center of Excellence, but it’s not right for everyone. If you have the ability to run at it and dedicate a team to develop the skills within your organization, then build your own Center of Excellence. What we are increasingly doing for our clients now is providing a Center of Excellence of dedicated people for your company. And we will handle the staffing, training, and provide our IP to them. The reason why we moved into this space is due to the large attrition of people who have currently been trained in RPA. As soon as they have that stamp or certification on their CV, they know what their potential is in the market. We’ve been able to retain all but 4 of our consultants so far. We invest a lot people, but I think the main reason is that we can provide career progression. There is always another opportunity for them in another role, which is hard to provide if you are staffing up a 2-3 person team in your business.
6. What is the cost to create and implement RPA solutions?
It is not priced on a business case model, so there is no practical overall price range to give. However, for enterprises, we see a common range for a single bot license that often falls between $2,000 to $15,000 per year. But this is just for licenses, whether done in-house or with third party - the license is only a small portion of the overall total cost of ownership. That bot can do the work of anything depending on how you deploy it. If you are just working normal business hours you might get a 1:1 ratio in terms of 1 bot doing 1 FTE worth of work. Or, if you are running it 24/7 you could possibly get 20 FTEs worth of work out of that. When we look at the business cases for our clients we commonly see 500%-800% ROIs over a period of 3-5 years.
7. How can we moderate the impact on employment levels where RPA can and will be implemented?
What I am really encouraged to see is that most of our clients are adopting RPA to back growth. It’s about how they handle more, with the same amount of resources. That has been the main focus, so we haven’t seen any clients that offer less jobs as a result of it so far. This might change as we go forward as we see more clients embrace the technology. I’m sure the analysts have seen different outcomes as well, looking at it from a global scale. There are two schools of thought. One is that we’ll always create more jobs and that we’ll always be creating new services and products in the market. The other is that it will have a negative effect on employment. So it depends on your perspective.