In their article “The Problem With AI Pilots” in the July edition of MIT Sloan Management Review, authors Thomas Davenport and Randy Bean pointed out that although “93% of respondents said their organizations were investing in AI initiatives” in a recent survey. they struggled to find companies that could share successful case studies. As they put it, “these were very large organizations spending goodly sums on AI and with a history of early adoption of other technologies.” Why the huge disconnect?
While some of this can be blamed on “lack of maturity” for AI technology, and organizational “change management” issues, a larger problem is that most organizations lack the clean, organized, data required to fuel successful AI deployments. Tamr Co-Founder and CEO Andy Palmer refers to this enterprise data mess as “random data salad” and warns that most AI projects fail because they are putting the AI Cart ahead of the Data Quality Horse. Andy explains:
“…the common thread unifying successful digital transformation programs and early wins from AI initiatives is that they start by unifying their core data from their many, many source systems. The advice I most often give to executives looking to compete on analytics is to make sure they aren’t putting the cart before the horse — get your data right first, capture early, impactful wins, increase DataOps maturity and THEN start integrating bright, shiny tech.”