The bots are here; but implementing them is not as simple as the hype might have us believe. Some say that Proof-of-Concepts (PoC) for Robotic Process Automation (RPA) can be dispensed with because the technology and methods – and the benefits – are now well proven. In DigiBlu, we disagree, but for different reasons. The implications and challenges of creating a digital workforce are profound and should be explored before embarking on full implementation.Software bots can be configured at warp speed compared to traditional IT developments and yet automations can still take longer, and be more complex, than is often claimed. Configuration is the easy bit, but success, we surmise, lies in being properly production-ready and in rapidly creating an effective Centre of Excellence and automation operating model for implementing the bots and managing the digital workforce. We share here some of the lessons we are learning from the experience of initiating RPA programmes:

  • The benefits of RPA are comprehensive, but you should be clear about your driving motivation (growth, innovation, agility, cost savings, compliance, accuracy?) because this will galvanise leadership, help to prioritise the selection of processes and inform the sequencing of the roadmap.
  • Be ready for the PoC. Access to your site, systems and networks should be made available to your Implementation Partner before starting and you need to consider system availability (planned downtimes) to set development timelines. Project roles & responsibilities should be clearly defined and accepted. These points may seem trivial but it can take longer to gain access than it takes to do the PoC design and configuration!
  • Define the solutions architecture, or at least the key principles, before deploying automations. We advise automating the easy tasks first – maybe 70% of the end-to-end process. This captures early, significant benefits, but leaves pockets of people handling more complex processing. You must decide which apps and systems to orchestrate and where to fill gaps using RPA. You may also reimagine candidate processes to further improve performance by leveraging RPA capabilities. The PoC demonstrates the power of automation and creates a ‘wow’ effect. Best then to pause, consider and align the key stakeholders and maybe do more design before rushing into production.
  • Because of the speed of automations, it can be difficult for business decision-making to keep up, at least the first time round. HR policy needs to change to allow staff ID’s to be issued to bots, for example, IT security and risk need to be convinced, budgets must be approved and items need to be placed on governance agendas. Implementing RPA requires strong leadership, effective governance and rapid decision-making if it is to succeed at pace.
  • Robotic automation may be “IT-lite” but the dependence on IT is high nonetheless. IT development and production environments need to be provisioned, time for testing put aside, test resource and data made available and servers need to be provisioned. While this adds time, IT must be brought along from the outset. Tempting as it may be, it would be counter-productive to create shadow IT environments because you must work within IT governance for RPA to be scaled.
  • Testing can become protracted because other IT changes may delay putting the bots into production. Contrary to received wisdom, the bots do go sick because they are sensitive to changes in the environment. This requires constant attention including collaboration with IT to anticipate changes and to troubleshoot the unexpected. Fortunately, it takes only minutes to reconfigure the bots and have them running again, but finding the problem in the first place can take much longer.

Most importantly, the potential changes to people and organisation should be considered from the outset. Initially, the bots may only automate some of the tasks that individuals perform and so people get shuffled around, released to meet demand elsewhere. Without careful management the benefits are not tracked and delivered. While RPA can be implemented tactically, through successive sprints, the end-state of an integrated human/digital workforce should be conceived early to achieve step-change performance improvements and not arrived at by chance having sown discord along the way. New roles and jobs will also be created. In DigiBlu, we’ve designed an operating model for virtual operations comprising leadership & governance, a design authority, RPA development and operations teams and a Centre of Excellence, and it is important that the implementation quickly builds these capabilities so that the organisation becomes self-sufficient in deploying automations and operating its digital workforce.

RPA is more than just another toolset and has the potential to fundamentally change the way in which operations are run and business services are delivered. Digital workforces create the opportunity for high-performance, agile, low-cost virtual operations, but getting there is a transformational journey and it helps to step off on the right foot by combining the benefits of rapid implementation, bottom-up as it were, with top-down strategy and planning.