Benefit Realization: The Horror Movie of RPA

In this article I will discuss why RPA projects are often failing to deliver and provide some simple advice on how to avoid the common pitfalls. Automation (RPA and beyond) is an extremely powerful tool, like many tools, in the wrong hands it can cause more damage than good but in the right hands you can create a masterpiece. 

As we see the exponential rise in the adoption of automation technologies, too we are seeing the frustration that many have with this promised land of low-code, simple cost saving tools. Building a robot isn’t hard, I agree with all the vendors that anyone can do it. But this isn’t the hard bit of the project, everything surrounding the development of a bot is. There is much more to an automation program than automation and as many of us have said for a number of years, Robotic Process Automation has very little to do with Robots.

If you are experiencing issues with an automation program not delivering benefits, or want to learn a few tips on how to avoid the common pitfalls, keep reading. This is in no way exhaustive; it is commentary on the most common issues.

Now, on with our fable.

The Plot:

How many times have we seen this movie? A company want to make cost savings in a back-office team, they attend a conference and hear about RPA and how it can cut costs in the back office with the simple application of low code robots. It seems too good to be true but there are case studies of how successful companies have been! Look at how many hours they have saved and what their FTE business case is! It seems at a good price point “One Robot can do the work of 3 humans, and they don’t make mistakes!” “It will cost less than $10k per year for a Robot worker”. All seems great then. Next up the vendor or a consultancy provide a high-level business case showing you under 12 month payback periods and circa 500% 3 year ROI. Any sensible person would take this right?! Indeed, many of you have.

Now, as with any good movie, comes the plot twist. You are 6 months into your program. Lets give everyone the benefit of the doubt and say there have been no major snags, no major system changes and nothing unexpected to impact the development roadmap, you are right where you should be (only around 25% of programs progress this smoothly in our observations). A review is done of the robots already in live and of the pipeline coming. The RPA CoE lead is jumping for joy with the ~10,000 hours robots have saved. You’re paying your bills to the license provider, any 3rd party development costs and the increasing cost of your own CoE, but that total payment is far lower than the 10,000 hours you have saved! But wait… mid year finance reviews and it shows that you have reduced profitability and your cost base has gone up. How can this be?!

The CIO/CTO are happy with the progress of the automation program and their KPI’s all look good. Number of Robots in production, infrastructure costs and hours returned to the business. The business are happy, they have motivated workers and things are being done faster but the CFO has noticed a spike in tech spend and no reduction in cost base. You may have seen incremental increases in productivity but nothing close to the spiraling tech costs. And here is where the RPA house of cards falls over…

What’s the issue?

We measure the wrong things, the tech is sold on the wrong KPIs and the business measure the wrong KPIs. This leads to a paradox of big cost benefits, per the RPA CoE lead and a cost increase per the CFO… how can both be true? Simple, hours back to the business is theoretical value, bottom line is tangible value. Hours back to the business is the best possible case saving, not the actual saving as many firms will have you believe. This erroneous way of reviewing your program is leading to the challenges that many automation programs are facing in not living up to expectations. Unless you know how to consume the benefit there is no value.

In many cases this is where the movie ends. Chalk it up to experience, we tried, we failed, move on. But it does not have to be this way. Automation can, and indeed has for hundreds of firms, led to very significant cost savings, improvements to productivity, increase in staff moral and a far stronger resilience to issues such as Covid-19 and other business impacting force majeure.

How do we get this right?

Firstly, there is no silver bullet. Anything worth doing is hard, don’t expect this to be simple either. There certainly are use cases that are very simple but a wide spread, scaled automation program is no mean feat, it requires tenacity, investment, leadership and experience as well as the willingness to try things, get them wrong and move on. There are a few tips to help you progress your programs.

1.      Top Down Leadership

The single biggest issues with any automation program I have helped recover has been the absence of top level leadership. Proof of Value, Proof of Concept both work in the micro scale with no executive sponsorship. A cross divisional, large scale automation program requires the sponsorship of executive management, preferably from the board. This way you can cut through regional/divisional politics and align everyone to the same goal. Without this leadership the chances of success are very slim.

2.      Focus on the tangible not the theoretical

Don’t buy into the smoke and mirrors of “hours back to the business” or “FTE savings”. These are directional at best, there is unlikely a single automation program in the world that has delivered to an “Hours saved” business case. This is simply due to the lack of consideration of Human Capital Change. 10,000 hours sounds great until you realize it is 30 minutes from each of your 20,000 employees. If it was 5 people at 2000 hours a year then that’s easy to act on, think about how you will extract and consume the hours. A benefit you cannot consume is not a benefit and remember, much like a gas expands to fill the space it is provided, work will expand to fill the time of your employees. Would you really be leaving work at 3pm on a Friday if I could automate 2 hours of your diary?

3.      Business Led, IT Governed.

The benefits of automation will be felt by the business, the impact of automation will occur in the business unit, yet many firms chose to make this an IT change. Think of RPA bots as digital workers, draw similarities to your human workforce. Now, if you needed to make a change in your billing department, say you are outsourcing the process, does HR lead this? In most cases no, but are they involved? Absolutely. The business take the decision to outsource the work, the business (with procurement or another team) draw up the description of services, run the sourcing activity, chose the vendor, implement the new process stack with the vendor. HR deal with any redundancies or retraining required. They certainly do not lead the outsource. Now, replace outsource with RPA and HR with IT. Why then, do so many firms drop RPA on IT and tell them to make savings in business units they rarely understand? IT absolutely should be involved. They should be the technical torchbearer as they will understand the IT Security, Info Sec and Infrastructure requirements better than anyone, but they should not lead this. They are the RPA police, not the RPA owners.

4.       This has nothing to do with robots. Human Change and Organizational Change are the keys to success.

This is an organizational change management program with a human capital change workstream alongside a technical implementation workstream. It requires a modification to your current operating model or even a complete redesign depending on the level of impact you are hoping to have with a digital workforce.

5.      See beyond RPA

RPA is easy to sell as it is simple to understand but it is not a panacea. Many other technologies such as OCR, ICR, NLP, elements of machine learning and specific “AI” instances will all be required to develop a true digital workforce. We talk about RPA being “hands work”, copy/paste, move something, manipulate something in a very mechanical “always-the-same” way. The rest is “head work”, listening, seeing, thinking, communicating. The Head guides the hands, without the head work the hands will only ever be capable of rudimentary, highly repetitive tasks, not the value adding capabilities most firms are now seeking from advanced technologies.

 6.      Measure twice, cut once

It is important to have a strategy and a plan. These do not need to be perfectly formed before you kick off but you should certainly know what you are trying to achieve and the major points on that journey. Detail can be filled in later but do not rely on blind luck. Invest in a strategy and in process targeting work. Ensure you budget for Change Management and Organizational redesign or your benefits will be eroded as you try to extract them from the project.

7.      Don’t hide in cupboards. Automation isn’t only for the back office.

Automation technologies are advancing at a rate much swifter than many realize. Gone are the days where Automation means scripting systems administration in IT or using RPA in finance. Automation today is large scale data and analytics driving intelligent machines to run core elements of your business and support your customer facing staff. It isn’t even as complicated as many people think, it just requires a bit of knowledge, some experience of the tools and a strong business acumen. We refer to this as the cognitive enterprise, using what is available around you to make better decisions, provide better customer service, better employee experience, and become a more profitable organization. Don’t hide these technologies in the back somewhere, embrace them and see what benefit they can have across the entire enterprise.

The Sequel: Automation Success & The Cognitive Enterprise

With a change of perspective, we can see things in very different light. For your next attempt at automation look at this from different viewpoints. How would IT see the problem, how about the business, how about Finance, HR etc? Do I need one or multiple technologies, ask not only “can I automate this?” but “Should I automate this?” and beware of naysayers, it is far too easy to fall into the trap of “this is hard” “it’s just a fad”. Kodak said the same about digital photography… whilst I urge you to plan and be realistic you must also be bold and tenacious.

Look to the middle and front offices for automation opportunities and avoid only looking for like-for-like replacements. Automation can fundamentally change how you deliver services, design thinking is a powerful tool to use in these situations, consider how the world would have looked if you had all of this capability when you first built the team you are reviewing or the product you are selling.

Consider your strategic approach well. If you are thinking Centralized RPA CoE, when was the last time you successfully built and scaled a CoE? Does your company culture work in this way? If you are considering Citizen Development and virtual assistants for everyone, how well consumed are simple systems such as Teams, Slack, Trello? If you don’t see adoption of simple tools why would people learn to use automation tech?

We do learn a lot from our past, we call it “Corporate Culture” but we can also change this and it should not on it’s own define our future.

I enjoy finishing with a quote, for this article I have chosen “audentes Fortuna adiuvat”  a Latin proverb meaning “fortune favours the bold”, often stylized as victory favours the brave. Whilst this was written in refence to battle, tech enabled change is no different. Be bold, be brave and you will stand a far higher chance of success.

If you have a specific question/challenge/point of interest I am always willing to discuss this, please feel free to reach out and contact me on LinkedIn or at my email address at the end of this article.