In the latest Internet of Things news for the insurance industry,
Yet another interesting thing about this technology is how it reflects a pattern Novarica repeatedly sees in the emerging technology space: Purpose-built solutions with a specific (but narrow) focus are easier to utilize in practice than general solutions that look to leverage an emerging technology for all purposes. Especially when a technology is new, general approaches leave insurers with the responsibility to figure out custom implementations that make sense and have value. But when a vendor steps in with a technology application that solves one key problem, everyone can immediately align on getting it into the real world. Essentially “purpose-built” devices are ready to go into the marketplace, while “all-purpose” devices are a platform that requires further implementation.
In the wearables space, we’ve seen several iterations of insurers looking for ways to leverage all-purpose devices. This includes Google Glass, most smartwatches, and other startups building devices to manage/monitor the entire scope of a work environment. Insurers attempting to utilize these solutions have typically backed off the projects due to costs, or the devices themselves struggled to find market momentum. On the other hand, IoT and wearables built with extremely narrow focus, like Kinetic for monitoring back injuries, or devices for gauging stream pressure, have done better with insurers because of the obvious cost/value equation and, just as importantly, the immediate use case requiring no additional implementation or planning.
A similar story exists with AI technology. There are a number of all-purpose AI solutions out there, the most notable being IBM’s Watson. So far, unfortunately, multiple insurer pilots with these AI platform technologies have been dropped, because training and building an insurance-specific function on top of the platform has proved costly, time-consuming, and requires strategic direction from both the insurer and vendor. However, purpose-built AI tools, like machine vision processing for assessing accident photos or aerial photography, or deep learning pre-built to scan news sites for company risk information, are getting picked up by insurers who might not even realize they are using artificial intelligence behind the black box. These solutions are more limited in ability and scope, but limits help with initial adoption.
This is not to say that the all-purpose platforms won’t eventually mature to the point that they supersede purpose-built applications. This has certainly been the case before. Many companies built or purchased purpose-built hand-held devices for activities like scanning and entering data, even though mobile phones were out there. But eventually mobile phones matured into the smartphones of today, and now most companies choose to build on top of the all-purpose smartphone platform rather than creating a secondary device. Certainly, IBM has the goal that Watson will one day be the AI platform of choice and that insurers (and other industries) will be able to turn it on and point it towards any problem. Likewise, more general-use IoT devices, such as a future iteration of Google Glass or a more mature smartwatch that can be leveraged for more enterprise functions, may end up competing with Kinetic.
However, until technologies mature and until the use cases are clear, purpose-built beats all-purpose.
This blog entry has been reprinted with permission
