After decades of promise and hype, artificial intelligence has finally become the defining technology of our era. Over the past few decades, the necessary ingredients have come together to propel AI beyond universities and research labs into the broader marketplace: powerful, inexpensive computer technologies; advanced algorithms, models, and systems; and, most important, huge amounts of all kinds of data.
The data-centric AI era started about 25 years ago when the explosive growth of the internet led to the emergence of what became known as big data, that is, very large amounts of digital data, including text, voice, and images. Digital data has continued to grow exponentially, and is now estimated to have reached around 150 zettabytes (1021 bytes).
“Collecting data, transforming it, and putting it to use is essential to remain competitive today,” wrote Thomas Davenport, Randy Bean, and Richard Wang in “CDO Agenda 2024: Navigating Data and Generative AI Frontiers,” a report sponsored by AWS for Data. “Virtually every organization is actively pursuing the goal of becoming more data-centric, enabling them to swiftly uncover and respond to valuable insights. With generative AI, the power and potential of organizational data is bigger than ever before. Organizations often appoint chief data officers (CDOs) to lead them toward data-driven success.”
The report is based on a global survey conducted in the summer of 2023 to understand how the role of the CDO and related titles like chief analytic officer and chief AI officer, (all hereafter referred to as CDOs), have been evolving over the past few years. The study received responses from 334 CDOs, and includes qualitative interviews with 12 leading CDOs. 66% of the responses came from the Americas, 18% from Europe, 10% from Asia Pacific, and 7% from the Middle East and Africa. The largest industry represented in the survey was financial services with 23% of respondents, followed by healthcare and life sciences (18%), government (10%), education (8%), and retail (5%). 20% of the CDOs reported directly to the CEO, 39% reported to other C-level executives, and 16% to non C-level executives.
Let me summarize some of the key findings in the study.
Creating visible value is still a key focus for many CDOs
“One of the key findings from our study is that CDOs need to show a visible value for their efforts — in part by emphasizing analytics and AI. As a relatively new C-level role, CDOs have more topics than ever to focus on, which makes it difficult to standardize the job description.”
Which of the following initiatives have you undertaken to bring more value to your organization? Over 50% of respondents said that they’re focused on a small set of key analytics or AI projects; nearly half have developed data literacy training for employees and have organized data, analytics, and AI councils; about 45% are focused on data management initiatives like improving their overall data infrastructure; 40% treat data like a strategic product within the organization; 40% are trying to ensure the success of data initiatives by publicizing their success within the company; and about 35% are focused on measuring the value achieved on each data project and on reusing analytics and AI functions within the organization.
Generative AI: The latest technology to create business value
“Many CDOs reported that they are responsible for generative AI. And while it is still early days for generative AI adoption, it’s clear that CDOs believe it is an important technology with an eventual profound influence on their organizations.”
How is your company addressing generative AI? About 25% said that they were experimenting with GenAI at an individual level; 21% said they were experimenting but with clear employee guidelines on GenAI usage; 19% were experimenting at the department or business unit level; 11% had organization-wide GenAI pilots; and only 6% have deployed one or more GenAI applications in production. 16% reported that they have not authorized the use of GenAI by their employees.
Which generative use cases is your company prioritizing? CDOs see a broad range of use cases: customer operations (45%), overall personal productivity (40%), software engineering (36%), marketing and sales (32%), and R&D (13%).
What is the biggest challenge for your organization in realizing the potential of generative AI? 45% of CDOs pointed to data quality and to finding the right use cases as the top two challenges, followed by establishing guardrails for responsible use (43%), data security and privacy (42%), GenAI skills (40%), data literacy and proficiency (36%), changing employee work processes (29%), and data infrastructure (25%).
How has your data environment changed to support or enable generative AI? No changes yet (57%) was the reply of the majority of respondents, followed by data integration and cleaning (25%), data to support GenAI use cases (18%), curating proprietary documents and text to train domain-specific applications (17%), partnerships with external data vendors to supply training data (11%), and purchase of domain-specific GenAI systems or services (6%).
CDOs’ jobs are still complex and challenging
“The CDO role is still relatively new among C-level executives and multifaceted. … While increasing numbers of companies are appointing CDOs, the job is still poorly understood for a variety of reasons.”
How well do you feel that the roles and responsibilities of the CDO role are understood compared to other C-level roles within your organization? The CDO job is still poorly understood was the answer of the vast majority of respondents (74%), only 8% said that the CDO role is better understood, and 18% said about the same.
Which of the following responsibilities are included in your job? The highest percentage was improving data quality (78%), followed by establishing clear and effective data governance (65%); data, analytics, and AI ethics (64%); building advanced analytics (61%) and business intelligence (59%) capabilities; improving the data infrastructure (58%); building AI capabilities (48%); GenAI strategy and execution (41%); data security and privacy (41%); and preventing cybersecurity breaches (15%).
To which of the following activities have you devoted 20 percent or more of your attention? Data governance (63%) was the top choice, followed by enabling new business initiatives based on data/analytics/AI (61%), data driven culture (56%), producing analytical or AI insights (41%), risk and regulatory compliance (33%), data ethics (28%), data monetization (16%), data breach prevention (14%), and fraud prevention (7%).
Greatest challenge for CDOs: creating a data-driven culture
“Consistent with other surveys, the challenges faced by the CDOs represented in this survey were mostly organizational and behavioral, rather than technological.”
What have been your greatest challenges in the CDO role? Difficulty in changing organizational behaviors and attitudes (41%) was the top challenge, followed by absence of data-driven cultures (35%), insufficient resources (32%), lack of data literacy (30%), unclear job definition (20%), lack of senior executive support (17%), and rapidly changing technologies (15%).
“The chief data officer role — increasingly combined with analytics and AI — is one of the most rapidly-changing jobs in business,” wrote Davenport, Bean, and Wang in conclusion. “As companies move toward digital transformation and data-driven decisions, it’s also one of the most central roles in that transformation.”
“A well-crafted data strategy — often listed as CDOs’ primary responsibility — is the foundation of generative AI success. CDOs will need to encourage experimentation, start with the right use cases, and treat organizational data responsibly to showcase visible business value with generative AI.”
Finally, the report lists ten keys to succeeding as a CDO:
- Constantly look for ways to add visible value to your organization.
- Add analytics and AI to the CDO portfolio whenever possible.
- Try to build coalitions and make other people successful in achieving their objectives.
- Encourage experimentation with generative AI, but try also to find strategic use cases for the technology.
- Don’t abandon existing data, analytics, and AI initiatives in favor of generative AI, but add it to the mix.
- Begin transforming and curating data, both structured and unstructured, to make it easier to succeed with generative AI.
- Adopt a common platform for data, analytics, and machine learning features for the organization to employ in its decision-making.
- Employ an “enablement” approach to achieving the data-related behaviors you desire, not a “governance” one.
- Take a use case by use case approach to improving data management.
- Strive to create a data-driven culture, but don’t force changes, and take them slowly.
Comments