10 Ways to derail an AI program (Harvard Business School)

By Harvard Business Review

Despite big investments, many organizations get disappointing results from their AI and analytics efforts. What makes programs go off track? Companies set themselves up to fail when:

  1. They lack a clear understanding of advanced analytics, staffing up with data scientists,engineers, and other key players without realizing how advanced and traditional analytics differ.
  2.  They don’t assess feasibility, business value, and time horizons, and launch pilots without thinking through how to balance short-term wins in the first year with longer-term payoffs.
  3. They have no strategy beyond a few use cases, tackling AI in an ad hoc way without considering the big picture opportunities and threats AI presents in their industry.
  4. They don’t clearly define key roles, because they don’t understand the tapestry of skill sets and tasks that a strong AI program requires
  5. They lack “translators,” or experts who can bridge the business and analytics realms by identifying high-value use cases, communicating business needs to tech experts, and generating buy-in with business users.engaged workforce positions a company’s digital initiatives for success.
  6. They isolate analytics from the business, rigidly centralizing it or locking it in poorly coordinated silos, rather than organizing it in ways that allow analytics and business experts to work closely together.
  7. They squander time and money on enterprise wide data cleaning instead of aligning data consolidation and cleanup with their most valuable use cases.
  8. They fully build out analytics platforms before identifying business cases,setting up architectures like data lakes without knowing what they’ll be needed for and often integrating platforms with legacy systems unnecessarily.
  9. They neglect to quantify analytics’ bottom line impact, lacking a performance management framework with clear metrics for tracking each initiative.
  10. They fail to focus on ethical, social, and regulatory implications, leaving themselves vulnerable to potential missteps when it comes to data acquisition and use, algorithmic bias, and other risks, and exposing themselves to social.

Read full article here…