Can Robotic Process Automation Deliver a Quick ROI?


Robotic Process Automation (RPA) is the application of automation technology that allows employees in a company to configure computer software or a “robot” to capture and interpret existing applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems.

Some people like to describe it as the new middleware with the ability to easily access and connect multiple systems as easily as a human workforce. There’s no need for heavy programming, or for lots of administrative staff and an IT team to manage repetitive, time-consuming tasks. What makes RPA unique is its ability to access everything through the desktop as a human operator would; read the internet, use passwords, make calculations, create /amend spreadsheets, make payments, read emails / PDFs etc, manipulate data, trigger responses and communicate with other digital systems.

92% of European Enterprise expects to adopt Robotic Process Automation by 2020. The survey of over 500 European businesses said their main drivers for adoption of RPA were improving customer experience and streamlining financial operations.

Gartner’s Chief of Research and Analyst, Fran Karamouzis has stated that “96% of clients are getting real value from RPA” and that the satisfaction rates were higher than she had seen for anything in her 17 years at Gartner.

With vendors quoting up to 50% costs savings for companies it would seem that RPA can not only deliver fantastic improvements in efficiencies in an organisation but it can also do it while bringing large cost savings.  But is this all too good to be true?  Is there real substance to the hype and, as with so many other new technologies, will the path to RPA adoption be littered with failed projects?  In this paper, we will cut through the hype to look at real efficiencies that might be achieved, and at the potential cost savings so you can make an informed decision on whether adopting RPA is right for your business.

First, we will look at where efficiencies might come from.

When people process large amounts of data manually, keying and rekeying into multiple systems, errors occur.  85% accuracy is a cross-industry benchmark for manual data entry.  If you apply the 1-10-100 rule of escalation as defined by George Labovitz and Yu Sang Chang in their 1992 publication “Making Quality Work: A Leadership Guide for the Results-Driven Manager” an error that would cost £1 to check at data entry point, will cost £10 to fix and £100 if left unfixed and goes onto affect customers.  In a system that allows only 1% error rate in manual data processing and input, those errors soon become a real cost to a business.

If we use the example of a financial company processing a loan application, where the company requires three years of personal and business tax returns with two guarantors and a business,  nine tax returns have to be entered. If these returns are 35 pages long with 25 fields per page, that is 875 fields at 100 characters per field, so 87,500 points of data for a single loan. Factor in an error rate of 1% for entering data from a paper tax return and the analyst is dealing with 875 incorrect data points.

Removing these errors removes cost and risk from a business. In this example, RPA can replace this keying of the tax returns, slashing the error rate and helping to remove the risk of an incorrect decision being made based on incorrect data.

With less time spent on the manual processes, human workers can improve the quality of their decisions as they focus on higher value tasks and the few exceptions that are still better handled by a human.

An organisation that needs to increase its speed of operations needs to improve efficiencies but even a 10% increase in efficiency from an employee does not change the simple fact that a UK based employee only works an average of 1,856 hours a year or just seventy-seven twenty-four hour days.  Reducing sick days, staff turnover etc will not have a significant impact on the speed of operations.

RPA can. And will.  One RPA software robot could work 24/7 365 days a year.  It will do the work of 4.75 people.  It won’t go off sick, take coffee breaks or chat with colleagues.

As an example, UK mobile communications provider Telefónica O2 deployed more than 160 robots to process between 400,000 and 500,000 transactions a month, delivering a return on investment of over 650 per cent over three years.

And physical staff need a physical location, which not only comes with a cost but also with risk.  If there is an issue in a location then manual processing may be slowed down or completely stopped.  While UK based employees face little more disruption than public transport disputes and delays, offshore staff may face larger challenges.  Natural disasters, geopolitical unrest and infrastructure outages can wreak havoc on offshore operations.

Adopting a system that is not location dependent such as RPA provides business continuity.

And if there are seasonal variations within a business, RPA can be scaled up and down quickly.  RPA is not like traditional IT projects.  In one organisation examined by Leslie Willcocks, Professor at London School of Economics (LSE) Department of Management, the return on investment for RPA was about 200 per cent in the first year and they could implement it within three months. The IT solution did the same thing but with a three-year payback period and it was going to take nine months to implement.

This ability to adapt and react is crucial. Every organisation the LES studied reported bigger workloads; an exponential amount of work to match the exponential increase in data—50 per cent more each year. And this, combined with increases in audit regulation and bureaucracy, puts a large burden on organisations who will be impacted if they fail to move fast enough.

So it is easy to see where cost savings could come from. By:

Reducing the number of back-office workers who spend up to 80 per cent of their time on repetitive manual tasks, lowering performance and motivation.

Freeing people to work on other projects as their current work can be completed without them or at least with them spending a lot less time doing it

Eliminating error

Increasing processing speed

Removing risk

But will these provide the 50% savings that are reported by some companies?

This, of course, is dependent on each company.  Many companies will have offshored much of their data processing and therefore the cost savings compared to a UK full-time employee will be lower.

To help understand the potential savings further we can look at the following example.

In this example, a company has already offshored a large part of their manual data input and processing with the majority of their transitional work being delivered from offshore locations.

  1. Cost reduction due to RPA implementation-

* 80 per cent of the total employee base replaced by RPA implementation. Of the 100 full-time employees, 80 can be replaced by RPA implementation

* A typical RPA license replaces 3 full-time employees, hence the number of RPA licenses required in above case is 27.

* Given that RPA is utilized to automate transactional processes, around 90% of the employees replaced by RPA are from offshore locations

* £672,00 per annum staff cost saving (based on £21,000 UK employee, £7,000 offshore employee (72 x £7,000 = £504,000, 8x £21,000= £168,000))

  1. Cost of implementing RPA

* One-time cost items, including consulting, implementation, configuring, and training fees, are estimated at £100,000

* One-time RPA deployment cost is amortized over the period of three years and added to the RPA license and operational cost to arrive at the total running cost of an RPA license (if Robots are 20, then the one-time cost per robot is: £100,000/27=£3,703 which over 3 year period is £1,234)

* The RPA license is £9,000 per annum and operational cost (of the person managing configuration and control) of RPA is £750 per annum

* Total cost of RPA implementation per annum in the above example= £206,180

This example has been simplified. There are multiple factors that need to be considered for the evaluation of adoption potential across different industries and business functions.

* Types of processes- Degree of presence of processes which are a) transaction/rule-based and b) consistent over time

* Volumes- Volume of transactional/rule-based work, and importance of horizontal business processes within different verticals

* Inefficiencies- Prevalence of issues such as the presence of disparate IT systems, issues related to transparency and compliance, and quality

* Nature of industry- Nature and the corresponding propensity of different industries towards the adoption of new technologies, e.g., the banking industry has a higher propensity than manufacturing industry, although all industries handle contracts and accounts and an ever-growing amount of data to collate, move and validate.

Understanding which processes can and cannot be replaced by RPA and drawing up a realistic project plan is essential to the success of any RPA rollout.  For many companies, this is the first time they will be doing this and to avoid making errors they should work with experienced consultants who can bring insight and expertise which will ultimately save thousands of pounds.  Companies who are already on the journey may wish to speed up deployments or expand them into other areas of the business and will require additional resource and skills.

James Aitman Director JifJaff

JifJaff is a specialised independent RPA consultancy. JifJaff can help you throughout your entire RPA project cycle. From initially accessing which processes will most benefit from automation and onto scoping the project. By helping you choose the right and relevant solutions and vendors for your business, we can reduce your overheads and increase your capacity, accuracy and profitability. We understand the difficulties you face and can advise on and implement an RPA system that will help your business embrace the future. We can design, deliver and manage your RPA solution. There are many RPA/AI systems, offering a multitude of functions. The breadth of choice can be overwhelming. Our consultants will bring their expertise to help you navigate the RPA minefield and advise you as to what is important and necessary for your company’s specific requirements, whether that is for a stand-alone project or companywide deployments.

Before you start your Robotic Process Automation project speak to JifJaff, we can save you money, remove risk and ensure the success of your project.