Robotic process automation (RPA) technology is currently the fastest-growing part of the enterprise software landscape. Does its appeal make more established approaches to automation obsolete? Let’s take a look.

Business process automation (BPA) technologies have been around for more than 20 years. An evolution of earlier document workflow technologies, these toolsets are used to build and manage digital systems that automate the co-ordination of work across teams, departments and organisations. Apart from efficiency and quality improvements, they can also bring transparency and agility to business processes.

For much of this time, BPA technologies have been quietly making businesses more efficient and effective — making a particular impact on organisations across the financial services, insurance, telecoms, government and healthcare sectors. However, until recently, the potential of BPA technologies has been limited — primarily because they were complicated to implement, expensive to procure, and because their implementation was typically tied to large-scale business change exercises.

Let’s contrast this with what’s been happening with RPA.

You’d be forgiven for thinking that RPA technology is a new thing — it’s been a headline-grabbing phenomenon only for the past 3 years or so. However, RPA technology has also been quietly in use, particularly in business process outsourcing (BPO) provider organisations and in some corners of the financial services industry, for much longer than that.

Whereas industry appetite for BPA implementations has remained modest, the industry appetite for RPA implementations has exploded. Organisations have been quick to appreciate the low initial costs commonly associated with RPA projects and the promise of fast returns, and have scrambled to apply RPA across finance, human resources, procurement, operations, customer service functions and beyond. The most successful RPA vendors are onboarding handfuls of new customers every day and are enjoying sky-high market valuations.

Can we just forget BPA and focus on RPA?

A question presents itself, then: Can RPA act as a cheaper and faster substitute for BPA? If RPA is quicker and cheaper, wouldn’t it be better to just use that technology instead?

The answer is that RPA and BPA are complementary, not interchangeable. Moreover, when RPA and BPA are implemented together, the results can be very powerful. Let’s quickly look at two of the ways that RPA and BPA are complementary:

  1. In BPA technology implementations, the focus for automation has primarily been the coordination and flow of work and the information that gets passed around as work is done. Individual tasks themselves often remain unautomated, particularly where those tasks need people to work with systems that are difficult to integrate. RPA implementations are primarily focused on the automation of precisely those tasks that BPA implementations haven’t typically touched, and only rarely do companies try to automate processes comprising more than a small handful of business tasks with RPA.
  2. Due to the low cost of acquisition and the speed with which companies can see results from automating individual tasks or short sequences of tasks, RPA projects often follow a “bottom-up” approach, looking at how to make quick, practical improvements to activities that cause frustration and delay. BPA projects, by contrast, have traditionally followed more of a “top-down” implementation approach, looking end-to-end at a target business process and systematically analysing how to make improvements that have an impact on overall process outcomes.

Friends, not enemies

RPA implementations can be quick and relatively cheap to execute, but RPA can be difficult to implement at scale because solving larger problems often becomes a question of automating or coordinating larger chunks of work activity. BPA technology helps here, because it can provide the necessary work coordination — acting as a conductor or orchestrator of tasks that are carried out by bots (as well as humans). RPA technology can also struggle where the systems to be automated express variation; bots can fail to execute tasks when unexpected conditions occur. Here again, BPA workflows can be used to blend humans and bots together, pulling in humans to help complete tasks that bots can’t.

More and more, organisations are employing both BPA and RPA technologies together — alongside artificial intelligence technologies as well — to improve the scope of tasks and decision-making that can be automated. It’s easy to see why.