Common, Critical Roadblocks to Enterprise AI Adoption

Data silos and management fiefdoms are common roadblocks to EAI adoption, and often appear in the early stages of strategic planning.

The C-Suite Paradox
A C-suite executive’s value is measured at least in part by the breadth and depth of their overarching oversight responsibilities. It’s natural for executives to want to carve out big pieces of pie, especially if those pieces are high-profile and potentially transformative. With Enterprise AI, though, there’s often no obvious “owner” at the C-suite level. The influence of this technology is – or should be – company-wide, and IT and analytics capabilities have already been decentralized. Some companies are creating new positions, such as Chief Data Officer or Chief AI Officer, to resolve this issue. Others are setting up “virtual organizations” internally to coordinate efforts, share lessons learned, and nurture the development of this critical core competency.

Data Silos
In a data-driven organization, data is a key company asset. A machine learning (ML) model requires that all of its data be co-located in a new kind of file storage system that no one department or division owns. There are a variety of tools to overcome the technical challenges of co-locating the data, and to move it efficiently between systems. Often, the more difficult obstacle is to convince employees and managers to share their data with other functional groups. This reluctance might come as a surprise until underlying causes and conditions are probed:

·       Data has surpassed knowledge on the power scale: human beings want to protect this asset and feel proprietary about “their” data.

·       Data typically cannot stand alone: a database is not as straightforward as it may seem, and data owners may consider it difficult or too time-consuming to explain collection practices, history, unique terminology and context required to correctly interpret the data, especially to “outsiders.”

·       Data quality may be compromised: bad habits, sloppy work, proprietary use, etc. is at risk of exposure.

Senior leaders who address these potential roadblocks in the early planning stages stand a better chance of integrating Enterprise successfully. EAI is a team sport that is best played by cross functional teams throughout every level in an organization. Data is a company asset, most valuable when shared company-wide.

Next Up: Strategic Planning for Enterprise AI