How to calculate the ROI of your RPA initiative?

Robotic Process Automation (RPA) should already be on your radar when it comes to business and process-oriented technology initiatives. Regarding its focus on automation and integration, it is typically a result-oriented productivity solution. In this sense, there are several paths and platforms available. But are these benefits worth your investment? The question is, how do you calculate the return on investment (ROI) for RPA projects?

The idea is to present the simplified vision for estimating the initial order of magnitude, without considering elements such as the internal rate of return, variability of monthly costs and other procedural elements that, if relevant, should compose a more structured calculation.

RPA Functions

Before diving into the details of ROI for RPA, it's important to understand the types of tasks that RPA is best suited to automate.

RPA takes on repetitive tasks that can consume a significant amount of time when performed by team members. These tasks are usually administrative or data management functions. These tasks are important to the day-to-day operation of a business, and mistakes in these tasks consume resources and can be costly.

Finding the best processes for implementing RPA

As a company adopts RPA technology, it systematizes the process of identifying automation opportunities based on their particularities. But, broadly speaking, what tasks should be prioritized for a company's initial RPA implementation?

1.Rule-based tasks

Clearly defined, rule-based tasks with few exceptions are ideal for starting RPA use. These tasks don't require much human decision making to accomplish.

2. Intensive work tasks

Tasks that require team members to spend more than an hour to complete are prime candidates for automation. Labor-intensive tasks are often time-consuming as well. RPA can dramatically reduce the amount of time these tasks consume your team.

3. Tasks with structured data

Inputs and outputs in data format with defined and searchable categories are ideal for automation. There are Artificial Intelligence (AI) tools that can help bring the efficiency of RPA to unstructured data, but for the most part, these tools won't be used until an RPA ROI analysis shows the value of automation.

How does this look like in the real world?

These types of tasks can be difficult to visualize. For example, we can mention how some functions of a Customer Relationship Management (CRM) platform are potentially automatable.

When using your CRM system to send a monthly action to your customer base, the data often needs to be reformatted, sorted and categorized for the email triggering service.

This is a pretty straightforward process when you only have a CRM, segmentation, and email to send, but if you run a business that sends marketing materials to multiple customers, you could easily have more than one full-time employee spending up to half their week extracting information from the CRM to lay the groundwork for campaigns. And this is just an example.

By using RPA to automate these tasks, your team will be free to use that time for more jobs that require human ingenuity that robots simply aren't capable of.

How to calculate ROI for RPA projects?

Now that we have some information about the types of tasks best suited to RPA, we can start looking at the ROI of RPA.

It's important to remember that there is more than one way to see how RPA benefits a business. There is the financial benefit, of course, but there are also more intangible benefits, such as employee motivation or the opportunity cost of errors avoided.

To get a complete picture of RPA's ROI, we must look at the financial costs, potential benefits, and how to factor in the human part of the equation.

What is the actual cost of implementation?

Because RPA software is tailored for each company that uses it, there is no single answer to how much the software will cost. It should be noted that RPA is relatively inexpensive compared to some solutions, but it will depend on the platform and licensing model.

While many companies have skilled IT staff, hiring RPA implementation companies can save upfront costs. An in-house IT specialist will need to take the time to learn how to design a scalable RPA system, design the process, and implement it.

While it might seem like it would cost less, it does mean that your specialist will forgo other roles for a longer period of time. By bringing in experts from outside your company, you streamline the process and save time and money.

In general, we can consider the annual licensing costs, the infrastructure, the costs of implementing, developing and sustaining the automations. Regarding the cost of licenses, it is worth evaluating the model and their potential occupancy so that the cost is distributed in more than one automation. Typically, considering the standard package from leading companies and the backlog of these packages, we consider the cost split between 4 or 5 initial automations. For this reason, it is even important that we start the automation journey with a list of processes and not a single work front.

Example

So, let's consider an initial package of licenses with a cost of $72,000.00 per year, divided between 5 automations. We are talking about a monthly cost associated with licenses of $1,200.00 per robot.

Regarding the cost of developments and cost of support, these can be applied directly to each automation. So, let's imagine a robot of moderate complexity with a one-time development cost of $40,000.00 and a maintenance cost of $1,000.00 per month.

We arrived, therefore, that this arbitrary robot would have a cost in YEAR 1, without considering the development period, of $5,333.34 per month. Let's keep this information.

Full-time equivalent (FTE) resources

An FTE (full time equivalent) represents the number of hours equivalent to a full-time employee. For example, for a journey of 8 hours a day, 5 days a week, 52 weeks a year, we are talking about an annual FTE of 2080 hours.

From this, it is important to know the total time consumed by manual activity, without automation. For example, if a process takes up 9360 hours per year, we are talking about 4.5 FTE.

The FTE can also be translated into values, considering the corresponding man-hours. Let us consider a full-time analyst with a salary of $2,000.00, thus resulting, with charges, in a monthly amount of approximately $4,280.00 per month.

ROI and Payback

Let's now imagine, after measurement and analysis, that the proposed automation generates a savings of 1.25 FTE, that is, 11700 hours per year. This represents a potential savings or monthly avoided cost of 1.25 x $4,280.00 = $5,350.00.

The ROI (Return on Investment) is calculated, considering a certain period, dividing (Earnings - Costs) by the Costs, in such a way that in the first year we have an approximate ROI of (5350 - 5334)/5334 = 0%. In the second year, we have the same gain but without the development cost, which represents a monthly cost of $2,400.00 monthly for automation.

So, considering two years (a good window for projects), we are talking about an ROI of ($128,400.00 - $92,808.00/ $92,808.00) = 38%.

Payback, in a simplified way, is the ratio between the investment (or cost) and the gains. Considering year 1, we can calculate approximate Payback of = R$64,008.00/R$64,200.00 = 1 . That is, this automation pays itself completely in 12 months.

What is good enough?

The first step is to establish these metrics and create a history. Many organizations do not have the system to evaluate or measure gains from this economic-financial point of view and waste the possibility of creating a knowledge base and generating internal benchmarks.

Getting closer to the area responsible for budget approval, management control and investment analysis is also essential. Defining and measuring gains must be a discipline of COE RPA with shared responsibility, especially in its validation, along with the business area. That is why it is important that this rationale is implemented as soon as possible with a documented record and with the approval of the main stakeholders.

Interpreting the fictitious result, we are saying that a moderately complex automation pays for itself in 12 months as long as it saves a little more than an FTE. This analysis and this criterion, contemplating low, medium and high complexity automations, is important to even direct the active search for themes, building a backlog that is potentially already oriented towards a relevant result for managers and senior management.

Of course, it is important to establish a compromise: the responsible area must realize these gains or demonstrate them in such a way as to validate the use case. This alignment needs to be formalized and should be the basis for the start of development.

For example, in a large client that we serve here in Brazil, a rationale was established primarily aimed at tangible gains, considering the costs and annual licensing model for a 12-month payback with demonstration of an effective reduction in the budget:

• Small Automation: Minimum earning of 1.0 FTE for two years
• Medium Automation: Minimum earning of 2.0 FTE for two years
• Complex Automation: Minimum earning of 4.0 FTE for two years.

Of course, these numbers are just an example and should not represent the reality of your organization. Obviously RPA generates greater gains than the objective result of FTE and this rationale can be explored in another article. In any case, it is worth reflecting on the importance of your organization carrying out a clear and transparent discussion on the measures of success of RPA technology from the actual survey of costs, exploration of licensing models and that the form of investment between the technology and business area. Also, formal alignment on how the results will be demonstrated so that the initiative is successful and with minimal frustration.

Gilberto Strafacci Neto

Country Manager at Practia in Brazil (www.practiaglobal.com.br) and Senior Partner at Setec Consulting Group (www.setecnet.com.br). Master Business Essentials CORe Program from Harvard Business School, MBA in Leadership and Innovation, Mechanical Engineer from Escola Politécnica da Universidade de S?o Paulo, Master Black Belt, Agile Coach, Design Thinker, Manager 3.0, Certified Six Sigma Master Black Belt from the American Society for Quality (ASQ) and Certified Scrum Master by the Scrum Alliance and Certified LEGO? SERIOUS PLAY? Facilitator (See PROFILE).