Artificial Intelligence & Machine Learning

AI Playbook for Chief Data and Analytics Officers: Part 1

Gartner Shares 10 Strategies to Take AI Solutions From Ideation to Disruption
AI Playbook for Chief Data and Analytics Officers: Part 1
Image: Shutterstock

In an era where artificial intelligence is reshaping the business landscape, Gartner unveiled a strategic playbook to guide chief data and analytics officers, or CDAOs, in their AI journey. The playbook, introduced by Sumit Agarwal, vice president analyst at Gartner, aims to help CDAOs achieve AI disruption and establish AI leadership for their organizations.

Gartner's 2023 CEO survey shows the need for a playbook, Agarwal said. The survey, which included 408 respondents, found that 21% of CEOs believe AI will significantly affect their industries in the next three years (see Image 1).

This C-level interest is putting pressure on CDAOs to implement AI solutions rapidly. "Some clients told me that their CEOs urged them to do something with AI in the next 90 days. One client even said their board has made AI a priority," Agarwal said.

Image 1 (Image: Gartner)

The Gartner playbook includes 10 strategies that define a structure for an organization's AI journey, the challenges encountered on that journey and how to address these challenges (see Image 2).

Image 2 (Image: Gartner)

Play No. 1: AI Ambition

The first play defines an organization's AI ambition using Gartner's AI Opportunity Radar. This four-axis spectrum helps organizations compartmentalize their ambitions into everyday AI and game-changing AI, as well as customer-facing and internal operations tasks (see Image 3).

The horizontal axis represents everyday AI and game-changing AI. Everyday AI helps improve productivity by enabling employees to work faster and more efficiently, freeing up time for other tasks. This includes using a generative AI tool, such as a copilot, to summarize a document, write an email or generate an image or a piece of code.

Game-changing AI focuses on disruptive innovations, and these AI tools can be used for creative tasks. Agarwal provided an example: An insurance provider may want to use AI to review pricing for various policies and claims to reduce losses. If the AI tool succeeds, it would be a game-changing move that will likely disrupt processes and take the business ahead of the competition.

The vertical axis of the AI Opportunity Radar represents customer-facing tasks and tasks involving internal operations. AI can be used for customer-facing tasks such as improving customer experience or billing. But AI tools can also be applied for internal tasks such as streamlining back-end processes within business units or for enhancing core capabilities such as R&D, supply chain and operations.

Agarwal said organizations should start their AI journeys with everyday AI before venturing into game-changing applications and consider use cases involving customer-facing tasks and internal operations.

Image 3 (Image: Gartner)

Play No. 2: AI Strategy

Business leaders face several questions when discussing their AI strategy: Is AI a priority? What kind of improvements can it introduce? How can they create competitive moats? These questions can be addressed by creating a compelling AI strategy that needs to be integral to your business strategy, Agarwal said (see Image 4).

"The strategy is your business enabler, and you need to collaborate with your business partners and business stakeholders. See what works for them, and think about the use cases that can enable and transform or disrupt the processes across your organization and various aspects of your business model," he said.

Organizations must take a holistic view when considering AI implementation, which includes using AI tools to improve productivity, create efficiencies, save time, grow faster, create new revenue streams and introduce automation. Agarwal cautioned against applying AI indiscriminately. Not everything needs to have a gen AI solution. For instance, a forecasting model that is based on predictive analytics and time series data may already be yielding accurate results.

Image 4 (Image: Gartner)

Play No. 3: Prioritize Use Cases

After defining the AI strategy and communicating it internally, internal business teams and development teams will likely generate use cases and proofs of concepts.

"POCs are often about technical feasibility, but one also needs to think about other dimensions such as value, cost, complexity and strategic alignment," Agarwal said. "This has to be thought about on three levels: Defend, Extend and Upend."

The Defend stage is the ideation stage where organizations are focused on improving competitive positioning through productivity and efficiency gains. The Extend stage involves creating differentiation within existing processes, such as implementing retrieval-augmented generation. At the final stage -Upend" - businesses work toward creating disruptive solutions or products to overcome the competition.

Image 5 (Image: Gartner)

Play No. 4: Measure Benefits and Costs

Organizations must consider multiple costs, even as they venture into new territory and attempt something new and unprecedented (see Image 6). For example, a life sciences company identifying new molecules may face high costs, but these costs can be justified by measuring benefits - such as productivity gains and cost savings - and communicating business value.

Image 6 (Image: Gartner)

Play No. 5: Fast-Cycle Innovation

The fifth play addresses the challenge of accelerating AI implementation. Agarwal acknowledged the increasing pressure from business leaders for shorter timelines. "We did a study and found that an average use case from start to finish takes about eight months. It takes six to nine months to train a machine learning model. But with gen AI, that timeline has become shorter," he said. To address this challenge, Gartner proposed a five-step approach for fast-cycle innovation (see Image 7):

  1. Ideate: Generate multiple use-case ideas. This step involves brainstorming and collecting a wide range of potential AI applications across the organization.
  2. Prioritize: Prioritize no more than a few. From the pool of ideas, select the ones that hold promise of generating the highest business value and focus on those.
  3. Build a team: Assemble a cross-functional team. Bring together experts from IT, business and security risk departments to develop and implement the chosen use cases.
  4. Design: Define a deployment approach and risk mitigation plan. Create a clear strategy for implementing the AI solution, including steps to address potential risks and challenges.
  5. Iterate: Deliver minimum functionality. Start with a minimal viable product, assess its performance and decide whether to stop the project, refine the solution or scale it up.
Image 7 (Image: Gartner)

Part 2 of this article will cover how Plays 6 to 10 help in managing the execution, risks and challenges of new AI projects.


About the Author

Brian Pereira

Brian Pereira

Sr. Director - Editorial, ISMG

Pereira has nearly three decades of journalism experience. He is the former editor of CHIP, InformationWeek and CISO MAG. He has also written for The Times of India and The Indian Express.




Around the Network

Our website uses cookies. Cookies enable us to provide the best experience possible and help us understand how visitors use our website. By browsing cio.inc, you agree to our use of cookies.