This article is published in collaboration with our key partner, Basico P/S.
An increasing number of companies decide to implement Robot Process Automation (RPA) – to optimize workflows and to automate consistent, rule-based, time-consuming, repetitive tasks and business processes. RPA can help to minimize errors and free resources for more business development tasks.
The growing interest in RPA naturally means that the number of RPA vendors is growing as well. Therefore, it may be challenging to find the right system that fits the processes and workflows of your company.
When assessing various RPA solutions for your company, you should, of course, always consider all relevant criteria and find the best suited RPA-tool based on how well the tool can meet your needs. There are many excellent RPA-tools on the market but none of them are perfect in all areas.
Here are five primary assessment criteria that we recommend focusing on when evaluating RPA tools:
The platform must be user-friendly when it comes to scripting, setup, installation, administration, etc. The easier it is, the faster you can train internal resources (not only IT-professionals) and share knowledge, thus achieve quicker results. The platform must be as flexible as possible, and it must be able to quickly solve automation tasks.
The RPA-platform must support the scalability of your company; it must be easy to utilize your licenses across the organization and to add extra licenses. It should be possible to start small with simple automation and scale in correlation with your experience towards a full-blown RPA automation setup. In addition, enabling local and simple desktop automation (RDA), the RPA-platform should include security features, the opportunity to generate activity logs, and functionality to perform centralized administration of the automation to make sure that you can securely grow your initiative towards governed RPA.
3. Democratic license structure
The prices for licenses must be easy to understand and transparent, and you should be able to foresee the ability to quickly achieve ROI. Both the initial investment and the future costs for implementation must be (fairly) easy to estimate based on a reasonable pricing model – you should be able to understand all the components of the platform and their individual prices. The price for making the software available to multiple users should be low or non-existent, it is essential to be able to include as many internal resources and stakeholders as possible without worrying about extra costs. The price transparency will make it easier to implement the platform in multiple locations throughout the business, thus increasing the returns.
4. Total costs
Besides the cost of licenses, you need to consider all surrounding costs. Internal training (is it easy to learn?), maintenance (can the users maintain the automation themselves or does it require IT expertise?), and development (how many people are able to learn to develop useful solutions and what does it require?). Furthermore, you must be prepared for the ongoing cost of maintaining and supporting the automation solutions. Systems change, processes change, therefore, your automation solutions will change over time. Robust and efficient error handling is a key feature and the RPA-tool must be able to handle different exceptions in the processes plus the troubleshooting process should be easy.
5. Vendor relationship
It is an advantage to partner with a vendor that is supporting your company and is open to discuss any business requests for system changes. If the vendor provides a strong support service (possibly online), well-written documentation and online training material, it strengthens the implementation process and later maintenance procedures.
In addition to the five assessment criteria, there may be situations where other criteria are relevant:
Some software vendors focus on promoting their long-term AI roadmap, which for some is very attractive, however, we recommend having concrete potential use cases where AI could help your company before actually using this as an actual assessment criterion. In most cases, companies are not mature enough (yet) to successfully implement AI.
If you are using an advanced Citrix infrastructure with published apps and similar features, you may consider testing the different RPA-tools in your environment to specifically see if any of them work well for you. Remote applications in general prevents RPA-tools to fully integrate with the target application, making automation more difficult. However, many companies simply apply minor adjustments to their Citrix environment or set up a parallel environment where the RPA-tool can thrive.
Software Scanning (Optical Character Recognition - OCR)
Some vendors promote their RPA-tool as being the best on the market on OCR. If you are working with a high volume of scanned documents and/or images, this might be a relevant assessment criterion for you to use in your evaluation, however, keep in mind that (almost) all RPA-tools can facilitate integrations to the leading solutions such as Google Cloud Vision to perform optimal OCR.
Foxtrot: Value-creating, intuitive and user-friendly
Foxtrot RPA is (probably) the most intuitive and user-friendly RPA-tool on the market. Foxtrot offers more than 375 actions, from simple mouse and keyboard actions to advanced API and programming features, and it is easy to set up automation in everything from Excel to, for example, SAP. One of the unique features of Foxtrot is the error handling that makes troubleshooting and maintenance extremely easy. This makes the process of scripting, testing, and implementing automation solutions user-friendly at a whole new level. Any errors can be corrected on the spot and the automation can be restarted from the same point of the error; just as automatic error handling is also an option.
The vast majority will be able to execute simple RPA tasks, as the user interface is intuitive and largely resembles already known software solutions. With a few hours of training, even "none-IT employees" can start automating their own tasks, which gives a completely different ownership to the concept of automation.