• by Ray Schultz , February 2, 2021

Most companies have a long way to go toward becoming mature users of AI and machine learning.

Only 17% have mature capabilities, according to a new study from Rackspace Technology: Are Organizations Succeeding at AI and ML?

Another 31% are moving from POC/pilot to an AI/ML solution in production, requiring significant organizational work. And 51% are only at the exploratory stage.

Worse, 34% report have attempted artificial intelligence R&D initiatives — and failed. The main problems were:

• Lack of data quality — 34%
• Lack of expertise within the organization — 34%
• Lack of productive-ready data — 31%
• Poorly conceived strategy — 31%
• Lack of integrated development environment — 29%
• Lack of investment in the right people — 28%

• Increased productivity — 33%
• Improved customer satisfaction — 32%
• Better streamlined processes — 30%
• Cost reduction in operations — 28%
• Increased level of innovation — 27%
• Increased understanding of your business and customers — 27%
• Improved decision making — 27%
• Enhanced performance/functionality of products — 27%
• Increased sales — 26%
• Faster time to profit—26%
• Faster time to insight — 25%

Here is how brands use AI and machine learning:

• Component of data analytics — 40%
• Driver of innovation in the company — 38%
• Applied to embedded systems — 35%
• Resource optimization — 34%
• Predictive maintenance/predictive failure—31%
• Create personalized customer journeys — 30%
• Reduce operational costs — 30%
• Drive new areas of monetization — 29%
• Product lifecycle management — 27%
• Optimization and testing—25%
• Automate marketing campaigns—14%

The top key performance indicators when measuring AI/ML performance are profit margins (52%), revenue growth (51%), data analysis (46%), and customer satisfaction/net promoter scores (46%).

Rackspace advises brands to clean up their data and data processes before diving into an AI/ML initiative, states Tolga Tarhan, chief technology officer at Rackspace Technology.