Data Leaders Can Create Business Value by Operationalizing DS and ML Programs

Our data continue to show increasing maturity overall for data science (DS) and machine learning (ML). User importance stands at a historic high across industries and roles. Industry importance sits near an all-time high. User perceptions of feature importance align well with industry support. Despite these indicators of generally increasing maturity, a sizeable gap still exists between user and industry perceptions about the importance of DS and ML.

Although data on DS / ML programs point to increasing longevity and ongoing adoption of DS and ML, about half of all programs and models lack sufficient maturity for organizations to operationalize DS and ML into more business-critical processes and systems.

Data leaders looking to advance their DS / ML programs should not focus primarily on the latest technology as a means to achieve their goal. Technology is not the problem: Our data show that industry support of DS and ML is robust and aligns well to user preferences and needs.

Instead, data leaders should focus more on accelerating production use of DS / ML models by solving the challenges associated with operationalizing DS and ML. These hurdles can include many key components of ModelOps and model life-cycle management, such as model-oversight responsibility, model governance, aligning DS / ML models to business outcomes, establishing KPIs for measuring model performance over time, and potentially revising or reengineering production and operational business processes (and their supporting systems) to include use of DS / ML models.

You do not have permission to access this document. Make sure you are logged in and/or please contact Danielle with further questions.

Rate this Research


Rated 0 out of 5
Very good0%