Commentary

Machine Learning's Slow Learners: Few Firms Lead In Delivering On A Data Strategy

Email marketers should forward the new MIT study, Building A High-Performance Data And AI organization,to their tech departments if they’re afraid to read it themselves.

It isn’t happy reading: Conducted by MIT Technology Review Insights in partnership with Databricks, it shows that only 13% of firms feel they excel in delivering on their data strategy. 

Among their challenges are: 

  • Data management platform does not easily scale — 44% 
  • Slow processing of large amounts of data — 39% 
  • Hard for cross-functional teams to collaborate on all analytics use cases — 29% 
  • High data duplication — 22%
  • Complex and fragment tools for ML — 20%

Moreover, only a few put themselves at the very top on a scale of one to 10 (10 being at the apex). 

advertisement

advertisement

10. — 2%

9. — 11% 

8. — 34% 

7. — 41% 

6. — 9%

5. — 1% 

4. — 2% 

3. — 0%

2. — 0%

1. — 0%

Only 13% in total call themselves high achievers, and 15% in North America.

The ratings per industry are: financial services (21%), government/public sector (20%), life sciences & health care (17%), oil and gas (17%, automotive & transportation (16%) and telecom (16%). 

That said, companies are trying to move forward. Cloud, once considered optional, is utilized by 63% in their data architecture. 

Companies worldwide hope to achieve these goals with enterprise data over the next two years:

  • Expand sales and services channels — 45%
  • Improve operational efficiency — 43%
  • Improvement innovation in reducing time to market — 42%
  • Improve maintenance of physical assets — 34%
  • Enter new product or service markets — 33% 
  • Improve ESG — 33%

To achieve those things, brands list the following as their most important enterprise data initiatives over the same period:

  • Improve data quality and processing — 48%
  • Increase adoption of cloud platforms — 43%
  • Enhance data analytics — 43%
  • Expand application of ML — 42%
  • Expand usage of all data (e.g., streaming and unstructured data — 36%

In North America, 50% hope to increase adoption of cloud platforms. But only 36% expect to enhance data analytics, perhaps because they already feel they are there. 

Next story loading loading..