Published on the 31/07/2019 | Written by Jonathan Cotton

A new report on robotic process automation looks at the opportunities and shortcomings of the emerging tech…

Robotic process automation (RPA) is the whitepaper topic du jour, and the latest from Juniper offers predictions of growth – but also a new focus on human/automation collaboration.

It all sounds good – lower costs, raised productivity, better customer experience. So what’s not to love?

The whitepaper Can RPA offer Better Customer Experience & Lower Costs? says that trinity of benefits means boomtime for RPA platforms and something of a productivity revolution across industries, with first in the queue being telcos and insurers.

Market pressure means telcos must offer cutting-edge, fast and affordable products, says Juniper. Enter RPA.

“With collaboration between human and virtual workers, employees can easily send tasks to a robot or vice-versa.”

“Controlling costs, managing data, developing new services, acquiring talent and increasing business agility are pain points that offer considerable challenges,” the report says. “RPA [is] specifically designed to assist companies in meeting and overcoming such challenges.”

How so? Brain-numbing activities like data entry, billing and ticket management are key activities for telecom companies, but they’re also time consuming and error prone, resulting in exceptionally high costs, says Juniper.

With investment in RPA, back-office activities can be automated, giving employees time to focus on complex and decision-making activities, such as customer services.

The power of automation extends to the front-office as well, such as customer-relationship building: “With collaboration between human and virtual workers, employees can easily send tasks to a robot or vice-versa. This leads to higher flexibility and better-quality services.”

Similarly, the insurance industry is facing its own market pressures, most acutely from a growing horde of insurtech competitors and an ever-shifting regulatory landscape, sometimes across multiple regulatory jurisdictions, making compliance complex and expensive.

“Regulations in the industry are often related to tracking, monitoring and reporting information promptly, which is especially challenging for legacy systems that were not built to provide such information,” explains the report.

“As automation allows precise data acquisition, analysis and delivery, it has become an efficient strategy for insurance companies.”

All this adds up to money in the bank for RPA platforms, says Juniper, with the total RPA platform revenue from insurance companies predicted to grow 244 percent – from US$184.2 million in 2019 to US$634 million in 2024.

But no discussion of software robots is complete without… chatbots. Any company with a significant degree of customer-facing activity will have to face the fact that there’s more than likely a chatbot rollout in their future, says Juniper. But take heart: While the robots will play a part in the helpdesk of tomorrow, that helpdesk will likely be manned by more than a few bits of code.

“When meaning depends on context and the relationship of the people involved in the conversation, chatbots can be very inefficient,” warns Juniper.

“Moreover, chatbot voice interaction is, in most cases, limited to scripted commands… Even with the use of machine learning, chatbots are only able to answer 20 percent of the questions they are asked.”

One possible solution, says the report, is combining the efficiency gains of automation tech with the nuanced decision-making of flesh and blood workers to create hyper-efficient man/machine customer services departments.

‘Pure’ chatbots “must learn all the information from the company to be able to address different types of questions”. Why not, therefore, provide robot-assisted conversational features to existing digital workers who are already performing tasks?

“These bots do not require a large input of information and they are more efficient than traditional chatbots as they are ‘specialists’ in the activities they perform.

“However, they are limited to those activities they already perform, while conventional chatbots are able to perform diverse tasks from different company sectors.”