Organizations Face Complex Governance Challenges with Analytics of Things

The Internet of Things and predictive analytics on huge stores of data are two technology storms beginning to collide. Well managed, they hold promises of improved operations and significant competitive advantage from new business models. Poorly managed, they hold threats of dangerous organizational risks.

At a recent Advanced Practices Council meeting, Michael Goul of Arizona State University referred to this convergence as the Analytics of Things (AoT) and noted the complexities of governing AoT applications that involve multiple ecosystem partners across industries, each of whom might own and use the predictive models and data stores critical to success. Imagine a scenario in which Robert, a 55 year old male diabetic, has an implanted wireless device that assesses his blood sugar and takes action. Using poor judgment, Robert eats a big slice of carrot cake for breakfast, sending his blood sugar level well above the normal range. The implanted device sends a message to Robert's smart phone that connects with a Smart City network interface that links to the Affordable Care Act Platform as well as his insurance company and physician. At the same time, the implanted device injects the right amount of insulin based on a predictive model. Later in the day, Robert falls and hurts his leg. The accelerometers in his phone recognize danger, causing his phone to send an emergency message to the closest emergency center, which sends an ambulance. An emergency tech arrives swiftly and starts first aid. When Robert's ambulance arrives at the hospital, his electronic health record is already available to the doctors on call. Think of all the ecosystem partners necessary to achieve such a well-orchestrated system.

Since contracts are the glue between ecosystem partners in such circumstances, Goul suggests paying particular attention to some key issues in these contracts.

Data exclusivity. If data ownership is to be exclusive, the contract should specify the specific entity that owns the data; otherwise, the contract should stipulate that data is non-exclusive because it is co-owned or open.

Co-mingling of predictive models. If an analytic model is created or derived through data that is co-owned, the contract should specify that the entities who co-own the data are co-mingled; otherwise, the contract should note that the entity that creates the model exclusively owns the model.

Data use after a partner leaves a partnership. The contract should specify who will own the data if a data owner leaves the partnership.

Analytic model use after a model owner leaves a partnership. The contract should specify who will own analytic models if the owner leaves the partnership.

Duration and fixed monetary exchange. The contract should specify both the duration of the contract as well as the basis for monetary exchange (e.g., data volume, number of customers).

Value-based compensation. The contract should specify how value derived from the relationship will divided among parties so that there are incentives for all stakeholders.

There are already lawsuits related to data ownership in these ecosystems, especially regarding indirect access to data.

#data #PredictiveAnalytics #analytics #analyticsofthings #IOT