Avoiding mistakes and exposing preconceptions: The importance of RPA pilots

Tags: RPA

AiiA Editor

Robots have been part of our lives for a long time. They have been constructing cars and working on farms for decades, yet their recent evolution has meant that many businesses are now looking at them with renewed interest. Robotic Process Automation (RPA) - software which allow enterprises to automate repetitive, mundane tasks once undertaken by humans, freeing them up to work on more strategic tasks – is the main reason for this pique in interest.

RPA is about delivering at the key moments, with efficiency and speed (having just binge-watched the Olympics it’s hard to get away from the vernacular!). And it’s due to this that robotics, and automation in general, is one of the most emergent trends in the shared services space right now and potentially, a successful RPA implementation could be imperative for upcoming innovation initiatives. 

Of the SSC’s that have implemented RPA, 36.4% believed that over 60% of their processes could be automated – a jump up from 3.1% who believe the same at the not implemented phase from SSON’s Dart report: RPA in Shared Services: Hype or Reality? In other words, companies often underestimate the effectiveness of RPA, and the importance of running a pilot program before going live.

As a major investment, anxiety from senior management teams is expected – there are always short-term challenges within the shared services industry which can take precedent over major overhauls. Those that neglect robotics, however, risk falling behind their competitors. Companies like Blockbuster and Kodak – once market leaders in their respective industries – failed to embrace new business models. And while the failings of these companies aren’t shared services specific, there’s a sense that history might repeat itself for those that discount RPA and AI.

If you are struggling to obtain management buy-in for robotics – or even if your company is steadfast about its RPA implementation and needs no further convincing – a pilot program is a good idea. Automating a few processes allows companies to prove the concept and test timeframes, which, ultimately, gives management a first-hand look at what RPA can do. In an interview with SSON’s Simon Barton, Vinoth Raman, Programme Manager at ING Bank, advised every company to start with a pilot, as it aids “readiness”. “If you want to implement RPA, you have the vendors to help you, but are you ready? If you want to implement RPA, are you prepared? Can you identify the process? That’s the biggest challenge; to make your organisation ready,” he says.

In addition to “readiness”, the advantages companies expect from RPA might not come to fruition. And according to Dart’s report, there was a clear difference between the perceived benefits of RPA, and the actual benefits, for SSCs. At the pilot phase, 52% agree that FTEs can focus more on value-added work as a major benefit of robotics; yet at the not implemented stage, the percentage was 35.3% – indicating that there is more to be gained than at first thought.  The opposite was found, however, when it came to RPA’s capacity to eliminate errors: prior to the testing phase 11.8% saw this as the main advantage, but by the time the testing process had begun, this had decreased to 4%.

This seems to indicate that a pilot program can highlight advantages, but also pitfalls. For example, Andrzej Kinastowski, Operations Delivery Strategist, Lufthansa Global Business Services, said that the pilot program allowed Lufthansa to spot that the RPA initiative wasn’t “scalable”. He said: “Our main problem with introducing RPA is our IT landscape.  Every separate Lufthansa company has its own ERP system, so we are looking at working with something like twenty SAP instances at this point, and they come with different processes and standardizations. This has been our main issue when it comes to automation: to design automations in a way that will serve more than one customer on one particular project. We hope RPA will help us design robots that can serve multiple customers.” The fact that the pilot program allowed Lufthansa to determine this shows its value.

In addition, if you plan to change your workforce’s position within the company – with the expectation that RPA will absolve them from certain duties – only for it to become evident that robotics is incapable of reaching the required productivity level, you risk irritating, if not alienating, your staff. By automating a few processes, however, you can accurately determine whether RPA will be able to take on certain tasks.

There’s also the burning question of whether you need a partner to help guide you through the process. Selecting your automation partner, or partners, is an important decision, and it’s imperative that the vendor is aligned to both your current and future needs – whether this is industry specific or through exact experience with certain processes. There are also a number of pricing models which companies must pay attention to, whether they’re license, value, or service based.


RPA implementation isn’t just about technology; organisational readiness is also imperative. The technology is out there – as are the vendors to help you on the way – but if a company isn’t prepared, diving straight into RPA could lead to organisational problems. The pilot program, in this instance, should revolve around raising awareness, and informing staff about their incoming “digital” counterparts.

Research indicates that pilot programs aren’t universally implemented at present, with just over half (56%) of those that have implemented RPA, running a pilot program, according to our Robotic Process Automation and Artificial Intelligence survey. Organisations might do well to rethink. A pilot – even if it involves just a few smaller, low risk tasks – can act as a blueprint for the wider implementation effort.