Savvy CDAOs will jump at the chance to use AI to deliver business value but must hone strategy and governance, as well.
- Gartner client? Log in for personalized search results.
AI for Data and Analytics: Focus on Where the Value Lies
Capture the promise of AI by making your data AI-ready
Data and analytics leaders must prove their organization’s data is ready to be used in an ever-growing number of AI initiatives, but vast differences exist between AI-ready data requirements and traditional data management. Download this roadmap to:
Ensure your data is ready for use in the specific AI initiatives you plan to pursue
Keep stakeholders aligned around what it really means for data to be AI-ready
Align AI for data and analytics through strategy, leadership and governance
The D&A leader’s success with AI hinges on delivering business value by integrating AI into the D&A strategy and prioritizing AI-ready data governance.
AI in D&A Strategy
AI Leadership
AI-Ready Data
Update the D&A strategy with an AI strategy, including attention to GenAI
AI initiatives are proliferating, often within a single organization in an uncoordinated and tactical manner. This is especially true of GenAI adoption. When one part of the organization is unaware of what another part is doing with AI, it can unwittingly create business risk and potentially destroy strategic business value. To prevent that, organizations need an AI strategy.
The point here is not to shut down proliferating AI initiatives but create a context for orchestrating them. Given that the chief data and analytics officer (CDAO) is already responsible for many enablers of AI — including the data and analytics and AI foundation, data governance and trust, data risk management, D&A ethics, analytical biases, data transparency, and parts of business-change-management through data and AI literacy — it follows that the D&A strategy must be extended to include the AI strategy.
Effectively framed, a data-driven AI strategy can inspire transformative thinking, orchestrate business outcomes, nurture informed decision making and define success. Yet too often, attempts to integrate AI into D&A strategies focus on high-level projects and projected budgets, while failing to identify how these AI initiatives will deliver value and by what metrics.
Avoid this trap with a more holistic view that connects AI adoption to business goals and ultimately business outcomes, as follows:
Align and map AI initiatives to the strategic business priorities of the organization; include metrics and KPIs for determining impact.
Engage enterprise stakeholders by speaking in the language of finance and business rather than the language of AI and D&A.
Understand the value levers and pain points of the organization and articulate how AI and D&A are delivering tangible value, linked directly to the business strategy.
GenAI as a special case within the D&A strategy
The sudden surge in interest around GenAI requires a specific strategic focus. Since this form of AI has widespread consumer application with a low barrier to adoption, teams can easily use it without organizational knowledge or oversight.
Even in the case of official GenAI initiatives, CDAOs are rarely the sole owners. Since you are responsible for AI capabilities such as data science and the underlying data needed, however, you can take a strategic role of providing perspective on where GenAI can add value today and where it cannot. Plus, you can provide leadership with critical guidance on what is needed for success in terms of data and trust, risk management and governance.
That enables the organization to focus on GenAI use cases that enable business value, and avoid getting distracted by hype and unreasonable expectations.
To enable that perspective, provide additional details on GenAI initiatives separate from traditional AI (predictive solutions) that include:
Closely tracking generative AI as a driver of both D&A strategy and of business outcomes
Identifying and prioritizing high-value GenAI use cases based on competitive impact, business value, urgency, cost and risk
Building proof of value and cost, then delivering via a portfolio of GenAI investments, measuring their impact, learning and correcting
Tracking emerging trends in GenAI to take advantage of the evolving value/cost advances as the market matures
In the struggle for AI ownership, CDAOs must show AI leadership
The sudden mainstream rise of generative AI is motivating ambitious CxOs to credibly argue that they should own AI in the organization. The CIO, CTO, chief digital officer, head of innovation, head of AI can all lay claim. Yet like AI itself, if explored in a reactive way, all this jockeying for ownership may be a distraction that slows the organization’s progress toward delivering value.
CDAOs need to be in the mix and have a voice in the AI conversation. You may not want to or need to own AI, but you need to be part of the organization’s AI leadership coalition. Your participation is essential, given that you are already responsible for many of the key enablers of AI — including the fundamental need for AI-ready data, data governance, and data and analytics skills and upskilling. You can also be a stabilizing force, if you show AI leadership by emphasizing discipline and practices that you are in a unique position to spearhead.
To keep the D&A organization central to the organization’s AI ambitions, CDAOs should take action in two key areas that differentiate you and your AI leadership:
Integrate AI into the D&A strategy with a focus on business value.
Seventy-four percent of CDAOs report that executive leadership has confidence in their D&A function, yet only 49% have established business-outcome-driven metrics that allow stakeholders to track D&A value. CDAOs may have been given a short-term honeymoon period, but that is over now. Without the ability to clearly tie D&A initiatives to value creation — including AI initiatives — CDAOs risk having their function dismantled and assimilated into the IT department, or into a data-heavy function.
Mature D&A governance as the key to business innovation and AI.
CDAOs increasingly recognize the importance of data governance for D&A success, and most organizations have successfully matured governance in recent years. For example, 82% of Gartner D&A survey respondents say they can identify the data assets needed for new D&A projects, and 80% commonly share a data asset across more than one use case. Gaps remain, however, as it relates to value-oriented KPIs for D&A governance, which only 46% of respondents have. D&A capabilities and delivery models must evolve to support business innovation and AI.
Fulfilling the AI value proposition requires high-quality, accessible data
D&A programs with highly mature D&A governance are most likely to have adopted data-driven innovations. This runs counter to the common perception among business stakeholders that governance disciplines can hinder innovation. On the contrary, a lack of governance prevents organizations from realizing value from their AI initiatives.
Gartner predicts that by 2027, 60% of organizations will fail to realize the expected value of their AI use cases due to incohesive ethical governance frameworks.
One clear gap is in establishing value-oriented KPIs for D&A governance policies, practices and procedures, which only 46% of organizations have. AI-ready data governance requirements are also different from traditional data management. Failing to recognize that will endanger the success of AI efforts.
To make data AI-ready, AI and D&A teams will need to be able to quickly identify data that is fit for use through three actions:
Align data to AI use cases. Every AI use case should describe what data it needs, depending on the AI technique used.
Qualify use. Ensure the data meets the requirements across the life cycle of the use case, from designing and training to operating an AI model.
Govern AI-ready data. Define the ongoing data governance requirements for the AI use case using parameters like data stewardship, data and AI standards and regulations, AI ethics requirements, and controlled inference and derivation.
The role of governance in AI-ready data
While D&A leaders strive to ensure its data governance enables AI-readiness, they must also respond to teams building AI that is data-centric — meaning, AI initiatives that prioritize engineering data as a path to building better AI systems rather than prioritizing refining and fine-tuning the algorithms or enhancing the code in AI models. For example, by 2025, synthetic data and transfer learning will reduce the volume of real data needed for AI by more than 50%.
Proactive CDAOs will drive data governance to enable AI-ready data through four enablers:
Data preparation and feature engineering. Data preparation mostly focuses around exploratory data analysis (EDA), cleansing and transformation to prepare high-quality structured datasets for feature extraction and engineering. Features add nuance or meaning to datasets for improving model performance and accuracy.
Data labeling and annotation. These critical, time-consuming and resource-intensive tasks involve adding metadata to unstructured data (images, text, videos and audio files) to identify features for AI development.
Synthetic data. Already common in computer modeling and simulations, synthetic data has emerged as an important resource for AI development. It’s projected to overshadow real data in the future due to its ability to retain the statistical and behavioral aspects of real datasets while optimizing scarce data, mitigating bias or preserving data privacy.
Data enrichment. Augment internal data with domain-specific data from external data sources. Data enrichment tools can gather third-party data from the internet (among other sources) and organize, clean and aggregate the data from disparate sources.
Gartner Data & Analytics Summit
Join the world's most important gathering of data analytics leaders along with Gartner experts to share valuable insights on technology, business and more.
Frequently asked question on AI for D&A
Can AI be used in data and analytics?
AI in data and analytics offers numerous benefits, including improved accuracy, predictive capabilities and enhanced decision making. By investing in quality data, choosing the right tools, developing talent and addressing challenges, businesses can leverage AI to gain a competitive edge.