Measuring RPA ROI – How to do it right?

Although RPA has been around for some years now, many companies are still at the early stages of their learning curve to start tapping into the technology’s full value promise. The first processes to be selected for automation are typically those that and tie up employees’ time from performing more valuable work and are highly repetitive, which, as a consequence, makes them error prone.

This is the right place to begin. However, as the use of RPA becomes more widespread within the organization more aspects need to be considered to drive the automation pipeline and to squeeze maximum benefits from the robotic workforce.

 A short guide to identifying suitable processes for RPA automation

Before we dive into RPA success measures & ROI calculations let’s run through some RPA basics. What processes CAN be automated with RPA? Here are our tips for identifying suitable processes for pure RPA automations:

1) Is the process rule-based?

Consider does the process consist of steps that are clearly defined – not up for interpretation – and are there a lot of exceptions to the rules? There might be too many process steps where a person needs to decide what to do next for the process to be worth automating. You may also find ways to improve the process so that the automated part would only be run a few times a month. These, and other reasons, may determine that the investment is just not worthwhile.

2) Does the process tie up a lot of time or valuable resources?

If performing some administrative process ties up a lot of the staff’s time from performing their main job it is clearly a candidate for automation. It is also clearly a process where savings can be easily calculated based on the freed-up hours. A rule of thumb is that automating a process should free up a minimum of 2 FTEs to be worthwhile.

3) Does the process have readable inputs?

RPA works on almost any application or environment but requires readable inputs – such as text-based data, user interface activities (like keyboard strokes and mouse clicks), Optical Character Recognition (OCR) fed data and green screen. However, inputs like non-OCR processed images or non-digital formats are not readable with RPA.

4) Is the data structured?

Structured data is comprised of clearly defined data types whose pattern makes them easily searchable. For example, detecting a specific term in an application data field signifies a certain action in the process. Unstructured data is – “everything else” – is data that is not as easily searchable, including formats like audio, video, and free-form writing.

Often the limitations of RPA related to non-readable inputs and unstructured data can be overcome by extending the technology with some type of AI component – Image and Voice RecognitionNatural Language Processing etc. In this article we won’t dive deeper into these tools, but you can learn more about them through the included links. You can also read about one of our most recent service offerings Document-as-a-Service (DaaS) here.

 RPA – The Queen of ROI

By nature, RPA is the Queen of ROI. It is fast to implement, doesn’t require changes to existing systems, and is fairly cheap. Especially when cloud RPA is delivered as a service it can be turned on and off like running water from a tap – with no fixed technology costs! In a turn-key service billing is directly based on usage, not capacity. The full investment consists of a development project and the (minute-based) service fee of running the robots.

The most basic RPA ROI is calculated from the below formula:

Cost of RPA Automation – (Hours Spent on Performing the Process Manually * Cost of Manual Labor)

Whether you implement RPA as a service or by licensing the technology, RPA ROI should be measured in weeks or months – never in years! A typical a pay-back time for an RPA projects is around 3 to 9 months. In a service model pay-back time is usually shorter than when licensing the technology due to lack of flexibility in adjusting the capacity to actual needs.

Other inhouse RPA costs are related to training or hiring experts, organizing the project management, and running the developed automations. Consider carefully your internal capabilities, the cost of developing these, and the key areas you should focus on in setting up and running your RPA program. For example, should you focus your resources to perfect reporting and analyzing your automation pipeline rather than begin a project to build an in-house team to develop and maintain automations?

If you would like to read more about the different organization models that can help you avoid RPA growth pains and provide you a solid foundation for scaling up check out our recent white paper about setting up a world-class RPA maintenance or book a meeting to discuss the topic with one of our experts.

Why is measuring key to success?

The use of RPA – as any significant business tool – should always be steered by company strategy. For example, if you aim to beat your competition by being best at customer service, prioritize processes for automation that support this objective. Simple, right? Well, it’s the first step.

RPA strategy should be derived from your general strategy, but also well defined. There needs to be clarity over responsibilities in running the program, and KPIs must be selected to fit your goals.

We have, seen many cases where developing adequate RPA success measures has been neglected. This is a problem for two reasons: It makes it difficult to steer the program effectively and establish the necessary support across the organization.

When trying to scale up and get different business-level stakeholders committed to the program, showing clear results is imperative. Not having systematic reporting in place is also problematic in a situation where people running the initiative change jobs and there isn’t clear evidence about the results to hand over to the next person deciding about the program’s faith.

Finally, results should advice the continuous development of your RPA program. Seeing what works makes it possible to make educated decisions about how, for example, your automation pipeline and prioritizing could be improved.

 What should you measure?

The selection of most suitable RPA KPIs depends on the strategic objectives of your automation program and the automated process itself. We have listed below some common RPA success measures to consider and to calculate into your RPA ROI.

FTE

This is the most obvious measure incorporated to RPA ROI calculations. The definition of FTE (full time equivalent) is the number of working hours that represents one full-time employee during a fixed time period, such as one month or one year. The cost of automating a process is compared to the cost of manually performing the process. FTE measures for calculating direct cost savings clearly apply to situations where the alternative to automation is, for example, hiring a large team of temporary workers to manually transfer data from an old system to a new one, or hiring staff from an external service provider to run a process outside regular office hours.

Hours-Back-to-Business

RPA automations rarely result to a situation where people become redundant. In fact, processes that are suitable for automation are typically ones that hold people back from performing their core job. For example, a doctor spending her time entering data to a system instead of treating her patients. Hours-Back-to-Business is measured by assigning a value to the hours freed up for more productive work or adjusting the FTE measure to reflect the value of returned work hours.

Tapping into the full value promise of freed-up time calls for Change Management skills. If a task is eliminated, what will be done with that time?

Reduced Quality Costs

This measure reflects by how much automating a process has reduced the need for corrections or eliminated mistakes in the process. Quality costs are costs related to preventing, detecting, and remediating issues in the process. If automation improves process quality, you need to know the cost of error. For example, getting rid of small regular typing errors in a buying process can generate savings by eliminating the need for cross-checking and reducing the number of filled slip orders.

Process Optimization & Lead Time

Process Lead Time refers to the time from the start of the work through the end of it in a process – which typically includes some waiting time. Automation can be used to optimize a process, for

example, by running preliminary actions at night or on weekends to reduce waste-time during office hours.

Speeding up Lead Time can also increase throughput, in some cases, even growing revenue! For example, if a bank is able to handle more loan applications with the help of RPA preprocessing around the clock.

Service Availability & Customer Satisfaction

One of the simplest ways to measure Service Availability is based on two numbers. You agree the amount of time that the service should be available over the reporting period. This is the agreed service time (AST). You measure any downtime (DT) during that period.

In a manual process, service-levels can fluctuate as result of, for example, key person dependency or request overload. Automation can improve the availability of service through freeing up capacity – resulting in higher throughput and shorter queues.

Automation may also enable a company to significantly extend its service hours – even around the clock. Over the recent years, digitalization has pushed customer service to evolve. People commonly expect online service to be available to them at any time of the day. Offering RPA enabled service 24/7, for example, could be a significant competitive advantage which’s business impact should be measured. Additionally, measuring the value of better service availability through customer experience metrics could come to question.

It can be difficult to incorporate Service Availability & Customer Satisfaction measures into your ROI calculations. But there are a few ways around this issue; You can calculate in the alternative cost of offering the same service-level without automation, or you can estimate the expected business value of improving Service Availability and/or Customer Satisfaction by X %.

Improved Business Agility

Improved Business Agility is a very important RPA success measure. The two things that commonly slow down business innovation are the lack of available human resources and technological rigidity. RPA can be used to add flexibility to both of these areas; free up human workers time and improve technological flexibility through, for example, better integration. Business Agility can be incorporated in your ROI calculations by measuring the amount of new strategic initiatives and assigning them a value.

Business Continuity Measures

Today, more than ever, the value of ensuring Business Continuity has been realized. Reducing human dependency in critical processes often equals reduced risks to Business Continuity. RPA is a great tool to achieve these goals. Calculate the cost of downtime and evaluate the risk. What is the value of reducing or eliminating these risks?

Employee Satisfaction

Employee Satisfaction is a soft measure but an important one. Overworked employees are often unhappy and if automation resolves their pain – measure it!

Employee Satisfaction measures could be incorporated in your RPA ROI through reduced staff turnover, better productivity, and improved employer image which makes it easier to attack new skills.

 There are other metrics that may suit your situation, but, from our experience, the above apply to most RPA ROI calculations, effective KPIs and automation pipeline prioritization criteria.

 If you would like to hear more about measuring from an RPA customer perspective, check out this webinar recording. In the webinar, our customer IF – a leading Nordic insurance company – talks about how they have perfected their RPA success measures and how systematic reporting has helped them steer the program.

About Digital Workforce

Digital Workforce is the leading company specializing in Robotic Process Automation  (RPA) and Intelligent Automation services on an industrial scale. Our intelligent digital workers automate knowledge work processes in large organzations freeing up the time of human employees for more valuable work. The deployment of digital workers requires no changes to the existing information systems. Digital Workforce was founded in the summer of 2015 and it currently employs over 240 IPA specialists in the US, the UK, Finland, Sweden, Norway, Denmark and Poland.

Digital Workforce offers a wide range of Intelligent Automation services that make it easy for your organization to design, build, run and maintain Digital Workers at scale. https://digitalworkforce.com

Ed Garabedian