In my previous blog on automation in BPO, I argued that, despite all the hype, the current implementation of robotics process automation (RPA) is little more than a labor arbitrage play, enabling the business to run with increased efficiency using existing technology and offshoring frameworks. So where does automation in BPO go from here?

RPA is essentially execution of repeatable, rule-based tasks which require little or no cognition or human expertise or human intervention (though RPA can be used to support agents within relatively hybrid tasks) by a bot mimicking human action. The bot operates enterprise software and applications through existing user interfaces based on pre-defined rules and inputs and is best suited to relatively heavy-duty transaction processing.

The next stage is to complement RPA with newer technologies such as AI where judgment-based tasks are starting to be supported with cognitive platforms. Examples of cognitive technologies include adaptive learning, speech recognition, natural language processing, and pattern identification algorithms. While RPA has typically been in full-swing for about two years and is currently reaching its peak of roll-out, cognitive technologies typically won’t reach wide-scale adoption for another few years. When they do, they promise to have most impact not in data-centric transactional processing activities but around unstructured sales and customer service content and processes.

O.K., so what about BPaaS?

Typically, BPaaS consists of a platform hosted by the vendor, ideally on a one-to-many basis, similarly to SaaS, complemented by operations personnel. These BPaaS platforms have been around for some time in areas such as finance & accounting in the form of systems of engagement surrounding core systems such as ERPs. In this context it is common for ERPs to be supplemented by specialist systems of engagement in support of processes such as order-to-cash and record-to-report. However, initially these implementations tended to be client-specific and one-to-one rather than one-to-many and true BPaaS.

Indeed, BPaaS remains a major trend within finance & accounting. While only start-ups and spin-offs seem likely to use BPaaS to support their full finance & accounting operations in the short-term, suppliers are increasingly spinning off individual towers such as accounts payable in BPaaS form.

However, where BPaaS is arguably coming into its own is in the form of systems of engagement (SoE) to tackle particular pain points and a number of vendors are developing systems of engagement that can be embedded with analytics to provide packaged, typically BPaaS, services. These systems of engagement and BPaaS services are sitting on top of systems of record in the form of, for example, core banking platforms or ERPs to tackle very specific pain points. Examples in the BFSI sector that are becoming increasingly common are BPaaS services around mortgage origination and KYC. Other areas currently being tackled by BPaaS include wealth management and triage around property & casualty underwriting.

In the same way that systems of engagement are currently required in the back-office to support ERPs, systems of engagement are starting to emerge in the front-office to provide a single view of the customer and to recommend “next best actions” both to agents and, increasingly, direct to consumers via digital channels. While automation in BPO in the form of BPaaS and systems of engagement is still largely centered in the back-office and beginning to be implemented in the middle-office, in future this approach and the more advanced forms of automation will be highly important in supporting sales and service in the front-office.

This change in BPaaS strategy from looking to replace core systems to providing systems of engagement surrounding the core systems offers a number of benefits, namely:

• It directly tackles the subject of improving specific KPIs and sub-processes within the organization in a focused and manageable manner, one at a time
• Secondly it offers the potential when combined with analytics to build an element of ongoing process improvement and learning directly into the sub-processes concerned
• Thirdly, it enables a modern digital interface to be implemented on top of the systems of record in support of agents, customers, suppliers, and employees
• Finally, it avoids the need to replace legacy systems on a wholesale basis and can be used to introduce transactional and gainshare pricing rather than licence and FTE-based pricing.

I started with the question “where does automation in BPO go from here?” In summary, BPO automation will only take a significant step forward when RPA becomes AI, and BPaaS emerges from the back-office to support sales and service in the front-office.