"These analytically driven findings create the second critically important consequence of the IoT: insurers can create feedback and control processes to command or request things to change their loss-related behavior and performance," he said.

According to the report, the Internet of Things has three, interdependent components: things with networked sensors, data stores, and analytic engines. In some cases, the IoT will provide more accurate data about various pricing factors that insurers are using, and in others, it can provide new kinds of data that could directly impact how insurers price.

The report reveals that the potential of the IoT to change the way insurers do business is immense, and explores the ways that this potential can be realized. Celent's report details how the IoT will give insurers first order data on a great many hazards and risks, and the ability to create feedback control processes to substantially reduce losses.

The IoT will change every part of the insurance value chain, including product design, pricing, underwriting, service and claims. The value to policyholders (in terms of reduced premiums and other costs of risks) and to insurers (in terms of reduced loss costs and expenses) must be balanced against the cost of retaining, maintaining and utilizing the IoT for success, the report reveals.

Over time, insurance premiums will decrease proportionately to decreases in losses. The report questions, though, will the insurance industry pursue an operating strategy that makes it smaller? And furthermore, if reduced premiums are materially large, will the industry accept a smaller role in the economy? Will it find alternative sources of revenue that build on the loss reduction role in the IoT?

Understanding the Internet of Things

To be a part of the IoT, things must interact, or have interacted, with network sensors that will measure and, in some cases, record and transmit the internal states or external status of the thing. According to Celent, the IoT things must be material objects, excluding intangibles such as intellectual property and reputation.

Internal states refer, for example, to a person's temperature and pulse, blood pressure, speed of a car or the stress on a bridge's welds. External status refers to the environment containing the thing, including room temperature, a speed limit or the weight of vehicles crossing the bridge.

Networked sensors send this data and information to data stores in the form of data, unstructured text, videos and other digital archival forms. Regardless of the form that the data takes, there will be massive amounts of data. Only the largest insurers will have the resources to store the data internally, and most of it will probably be stored in a cloud.

For insurers, what is critical about this data is that it can provide a much more accurate picture of the exposures, hazards and risks of what is being insured.

Taking into account the four categories connected to the IoT—living things, moving things, stationary things and transmitting things, analysis of the IoT can provide insurers with data that reveals what is possible today, what is likely to be possible over the next three to five years and what insurers may or may not be allowed to do, depending on legal and regulatory frameworks and policyholder preferences.

To an extent, this data is already available to insurers. The IoT, however, could bring its new and better time series data, revealing, for instance, how fast a vehicle was driven over given roads with given speed limits. Insurers' analytic engines can parse the data to create better products, pricing and underwriting guidelines.

IoT's potential to change the way insurers do business is massive, but there are three operational contingencies, including the development of efficient and useful sensor technology and sensor networks, the deployment of networked sensors on and in many insured things and Insurers' Big Data analytic capabilities to store, model and create products, pricing algorithms and underwriting decisions.

In order for the IoT to impact losses and premiums, there needs to be a direct relationship between what is being sensed and hazards and risks, and things have to be willing or able to change their behavior and performance.

While the IoT has potential to make the world safer, it does not come without dangers. The IoT certainly has the capacity for making the world safer. But it also has the capacity to make it a more dangerous and intrusive place. Sensors, analyses and models, and commands and requests are the three elements that are subject to failure, Celent claims.

Impact on the Value Chain

However, IoT could add value to the insurance chain, once a number of elements are addressed. In product design, for example, the terms of an insurance policy will have to address the presence and operating conditions of sensors. In pricing, some of the traditional pricing elements will need to be replaced by the data elements found to be the most powerful predictors of losses by the insurers' analyses and models.

In underwriting, scores and decisions will incorporate new elements that reflect the data sets built during prior policy periods. Decisions also may draw on a broad range of data, such as images and videos.

Celent predicts that policyholder service will dramatically change with the addition of the responsibility for the operation of the feedback control process. According to the report, insurers will need to learn entirely new skill sets and the best ways to induce people to reduce the likelihood of loss.

Claims operations will use new elements to understand causation and responsibility, such as videos of vehicle accidents, for example. Fraud mitigation tools will also use broader and better data algorithms.

In the history of P&C insurance, the primary focus has been understanding risk, pricing risk and paying losses. A secondary focus has been on preventing losses. However, for these purposes, the industry has often turned to second-order data.

"The IoT will give insurers first order data on a great many hazards and risks, and the ability to create feedback control processes to substantially reduce losses," Light said.

If the insurance model is changing, as the report asserts, alternate IoT revenue sources could include some activities that fall within the traditional insurance service of loss control. It is a matter of turning to analyzing the policyholders' hazards and risks, providing actionable recommendations for minimizing losses, creating new products, information services and gamification in moving forward.