Article | 18.10.2022
How to start with RPA: A Guide to automation adoption
Many still have might have the perception that RPA requires a long-term commitment, or purchase of expensive licences and technological expertise to get started. However, robotics can also be tried with a low threshold and it is suitable for budgets of all sizes. An experienced automation partner facilitates the implementation of automation at different stages of automation maturity.
Perceptions of expensive and lengthy automation projects still persist. And not without reason, as some studies show that 50% of automation projects fail, and many companies get stuck with software robotics after the first robot is built. Indeed, a failed automation project or failure to scale up automation can make robotics seem expensive compared to the investment.
However, an experienced automation partner will identify which processes are worth automating, what errors might occur and how to prevent them. A skilled partner will also tie the automation project to business-level KPIs, measure automation results over time, and develop an operating model that allows scaling the automations. You can read more here about how we ensure that the automation project succeed.
What does it cost to start using RPA?
You can experiment with software robotics with budgets of different sizes. Costs depend largely on the number of automations and whether RPA is bought as a service or whether the company wants to buy the licences for the technology itself. At the lowest end, automation can be tested with an annual budget of €10,000 to €15,000 when bought as a service. If a company buys the licenses for the technology itself, the license costs range from €30 000 to €100 000 per year.
When estimating the cost of automation, it is worth examining how the annual costs behave: when costs are spread over several automations, the unit cost is always lower. Fixed operational costs associated with running the technology platform, such as licenses and resources, are roughly the same for one automation as they would be for five. Therefore it’s essential to scale up the automation, and get the whole organisation on board, so that the benefits of automation are truly realised and become more profitable.
Before starting with automation, you should make a business case calculation of RPA’s ROI
We also recommend looking at the cost of the robot and its business value over the whole lifetime. Often automations run for 2-5 years, so the initial development cost is a relatively small cost. But above all, the value that the robot delivers needs to be understood and it is therefore important to measure the value that automation of knowledge work can deliver. For example, if implementing a software robot can avoid hiring a new employee, this will already save €55,000 annual cost per person. Or if an automation succeeds to save 5 million in the supply chain by reducing goods-in-transit, the cost of RPA is a very small factor.
Once we understand what value or values are being driven by automation, the next step is to find out in what timeframe the cost will payback itself, i.e. what is the ROI of RPA. For example, when we built a robot to provide an SMS payment reminder service for our client Y-Säätiö, the robot investment paid for itself in only a few months, just by reducing rent receivables. You can read the full story and about the other results achieved through the automation here.
Our role as an automation partner is to help you identify the life cycle cost and payback period of your automation. The automation must realise the business objectives set for it, such as saving working hours or speeding up process throughput. An experienced automation partner helps you calculate the return on your investment in RPA within the first year. Sisua Digital has an analytical methodology for this and this is typically how we start with our customers.
Is it more profitable to buy RPA as a service or pay for the licenses yourself?
Buying RPA as a service (Raas: Robotics as a Service) is often the easiest and most cost-effective solution to get the fastest results. The total cost of the automation is often the cheapest in this option. When automation is purchased as a service, the cost of RPA is also transferred from Capex to Opex, so the monthly payments accumulate at the same rate as the benefits, which also speeds up the payback period.
Most companies have moved to this service model in the last years, and according to Forbes, for example, the RaaS market is growing rapidly. The model chosen by Sisua’s clients is RaaS – you can read about our client’s experience with RaaS here
Benefits of the RPA as a Service model
- A flexible way to use RPA
- Scaling is easy and cost-effective
- The “trial cost” of automation is much lower than if you had to buy the technology yourself
- SMEs can experiment with RPA without a large upfront investment
- A service partner can provide both licenses and maintenance services at a lower cost than the licenses would cost
- The expertise provided by the partner improves the performance of the automation and the partner ensures that the necessary updates and maintenances are carried out
- When the partner provides the licenses and the expertise, the company can focus on its core business
An alternative to Raas service is, for example, an on-premise model, where the technology runs on the customer’s own servers and requires its own licences. Alternatively, the technology can be hosted in a private cloud. Such solutions often require an initial investment of around €100,000.
Some technology providers may also offer the technology as a cloud service, where it comes from the technology provider’s cloud. Often these are in a price range close to that of buying your own licenses, and do not include any specialist work.
A tip for choosing the RPA technology
We do not recommend committing to any marginal technology, whereby the company is married to a single supplier. So you should avoid locking yourself into a single service. In particular, when buying on-premise technology, the company is tied to that particular technology, whereas when buying as a service, the company can easily disengage from the technology.
How quickly can a software robot be deployed?
At best, a robot can be up and running in just a few weeks. But the delivery time depends very much on how data-intensive the process is, and what format the data is in and how much test data is available. Often it takes just a few weeks to validate and see what the robot could look like in production.
However, as all processes are individual, it is not possible to give an exact timeframe. Automation can be put into production very quickly, or you may have to wait 3 months for test data before you can even start planning how to do it.
In those cases where RPA has already been identified as the best solution, development is often rapid. When it is already reasonably identified what the robot will be used for, results can be yielded faster.
What does implementing RPA require from a company?
To put it simply, if you buy RPA as a service, then the implementation does not require harldy anything from companies in advance. The company doesn’t need to know which technology to choose or other details, but the service partner will help spar over what could be the right solution for that particular company, and what would be a good automation target and which representatives from the company the automation project should start with.
So there is no need to have an RPA function or a CoE (Centre of Excellence) in place, as the service partner will get the project off the ground. But it would be good to have an owner, or a steering group, on the customer side. RPA needs a sponsor, a decision maker and a lead user. Often, it is the experts in the process or area who are these end-users, who best understand the process that will be transferred to the robot. If there are delays in an automation project, it is often because the company does not have the staff in place.
How do you proceed with RPA if you don't have an automation target in mind beforehand?
In such cases, the potential of automation should be assessed by gathering information from experts on bottlenecks or routine tasks by conducting workshops in different functions or teams in the company. The idea of workshops is to surface tasks and processes that involve a lot of human effort or error-prone work and where metrics about business improvement can be found. This option is useful when you want to first explore the potential and impact of robotics before moving forward.
If, on the other hand, the company has already identified which work routines it wants to move away from specialists, then a lighter analysis is made of how best to solve the problem.
How to increase automation know-how in your company?
Internal automation skills can be increased through training and citizen development programmes. As a partner, we spar with the customer and make it easy to build automation capabilities from the customer’s end and work as a team with the customer. Our service platform allows the customer to develop automation themselves and we provide training for staff.
How to present the new technology in the company?
We recommend that you carry out staff briefings before starting with robotics. Typically, this service will move employees’ work to a machine, so it’s really important to communicate what RPA is like, and how robots will become like digital assistants to employees. It will help to remove concerns that robots will not take jobs. The aim is to get employees to see how robotics works and to experience that it is a good thing and not a threat. Then you can also start involving more employees about what other routines could be automated.
Author: Paiju Koivula
Interested in the potential of automation?
Book a free consultation with us
Want to know how automation could benefit your company? Tell us about your automation needs and let’s find the right solution for your company together.
See what kind of work we have done for our customers.
GET IN TOUCH
Learn more about how we can help your company.
Read about what’s happening in the field of robotics and AI.