In Growth in the Machine, Capgemini’s July 2018 report on the state of intelligent automation in the financial services industry, 33% of surveyed insurance executives said their firm had realized top-line growth of 2 to 5% through intelligent automation solutions. Moreover, 65% of insurers said they had improved customer satisfaction by more than 60% through intelligent automation.
However, only 8% of the insurers polled said they had implemented intelligent automation projects at full scale.
But, progress is sluggish
Challenges that prevent insurers from moving beyond proof of concept or the pilot stage are three-pronged and related to business, technology, and talent.
Business: Identifying the right automation process is the first step, but the inability to establish a business case is a concern that, more often than not, elicits an anemic response from senior leadership. Also, a lack of coordination and consensus among stakeholders and business units makes it difficult to reach agreement about which processes to be optimized or the support roles needed.
Technology: AI-based systems require the right kind of data in sufficient volumes. However, the quality and usability of data remain an issue in the insurance industry. The lack of an appropriate data management strategy hampers automation progress.
System integration is crucial for the seamless operation of an automation solution. It can be challenging, however. For an automation program to run successfully, it must access all the source data from multiple systems. Getting and maintaining that access is made more complicated by technology integration, change control, and other risk and control issues.
Cybersecurity and data privacy are also major technological concerns. Since the EU’s General Data Protection Regulation (GDPR) came into force earlier this year, the importance of data privacy and security has surged for firms that handle personally identifiable information (PII). Insurers must now safeguard PII, such as names and addresses, to prevent significant penalties arising from its misuse. Therefore, challenges related to access rights and firewalls must be tackled to enable full-scale implementations.
Talent: Considering that some employees may view automation as a potential job threat, it may be necessary to overcome internal resistance before launching an intelligent automation initiative. Since scaling up requires a profound understanding of emerging technologies and their implementation, recruiting and training automation talent are also among the issues insurers face.
Considering these challenges, it isn’t surprising that more than half of insurers have yet to move past the pilot phase for RPA bots, and the number is even higher for intelligent technologies such as AI or machine/deep learning, according to results from the World Insurance Report 2018 from Capgemini and Efma.
The need to clear obstacles is strategically critical.
In my next blog, I will share a roadmap to help insurance organizations determine their current state of automation and consider how to proceed pragmatically to scaling it up. Meanwhile, I would love to hear your thoughts. To discuss this further, connect with me via my profile or on social media.
 Insurance Business Canada, “Insurers say 60–80% of data they collect is ‘not accessible,’” Bethan Moorcraft, March 19, 2018, https://www.insurancebusinessmag.com/ca/news/breaking-news/insurers-say-6080-of-data-they-collect-is-not-accessible-95255.aspx