Developing a Generative AI Strategy to Drive Value in Your Business

Developing a Generative AI Strategy to Drive Value in Your Business

AI strategy: From setting AI vision to executing value-driving AI initiatives

Generative AI (GenAI) is one type of AI that executives suddenly want to try in their business, but to capture its value and manage risk in a sustainable way, executives need a sound, holistic and achievable AI strategy.

Consider the four key elements of any AI strategy (below) and download the GenAI planning workbook to:

  • Set GenAI goals, benefits and success metrics
  • Tie your GenAI vision to business impact
  • Assess and mitigate major AI risks
  • Prioritize GenAI initiatives

4 AI strategy pillars keep you focused on driving business impact

Building an AI strategy inclusive of GenAI requires a rigorous approach — from developing a business-driven vision to planning which initiatives to adopt and why.

1. AI Vision: Identify the strategic opportunities of generative or other AI

Generative AI (GenAI) has become a significant focus in the business world, offering transformative potential similar to past innovations like the internet and electricity. Its impact on enterprises spans various aspects:

  1. AI Maturity in Organizations: According to Gartner research, only about 10% of organizations experimenting with AI are considered mature in its deployment across multiple business units and processes. These mature AI organizations provide valuable insights for those looking to adopt GenAI.
  2. Role in Enterprise Growth: GenAI is poised to support enterprise ambitions and drive stronger results. Its deployment can become a competitive advantage and differentiator, especially in automating repetitive tasks and generating new insights and innovations through predictive analytics, machine learning, and other AI methods.
  3. Impact on Shareholder Value: GenAI can create disruptive opportunities to drive enterprise goals. For example:
  • Increasing Revenue: In industries like pharma, healthcare, and manufacturing (including CPG, food and beverages, chemicals, and materials science), AI can accelerate the creation of new products. This ranges from new drugs to less-toxic household cleaners, novel flavors, fragrances, new alloys, and improved medical diagnoses.
  • Enhancing Customer Engagement: By disrupting traditional value chains and business models, GenAI enables direct content distribution to consumers, bypassing intermediaries like publishers and distributors, thus improving customer engagement.
  1. Cost Reduction and Productivity Improvement: GenAI capabilities can simplify processes and accelerate outcomes. This includes augmenting human efforts (e.g., content summarization, classification), generating software code, optimizing chatbot performance, and utilizing previously untapped data.

Overall, Generative AI represents a paradigm shift in how businesses operate, innovate, and interact with customers, with the potential to reshape industries and redefine competitive landscapes.

Decide how to measure the success of AI

A recent Gartner survey of more than 600 organizations that have deployed AI shows those with the widest, deepest and longest experience with AI do not measure success by project volume, tasks completed or output. Instead, they:

  • Emphasis on Business Metrics Over Financial Metrics: Successful AI deployments focus on business metrics rather than just financial outcomes. These metrics are tailored to specific use cases, employing attribution models and ad hoc measures that align with the unique aspects of each application.
  • Benchmarking Practices: These organizations engage in both internal and external benchmarking. Internal benchmarking helps in assessing progress against past performance and organizational goals, while external benchmarking provides insights into industry standards and competitive positioning.
  • Early Identification and Consistent Measurement of Metrics: Identifying relevant metrics early in the AI project lifecycle and consistently measuring the success of AI use cases is a common practice. This approach enables timely adjustments and ensures alignment with strategic objectives.
  • Diverse Business Metrics: Various business metrics are considered, including:
  • Business Growth: Metrics like cross-selling potential, price increases, demand estimation, and monetization of new assets.
  • Customer Success: Including retention measures, customer satisfaction, and share of customer wallet.
  • Cost-Efficiency: Metrics related to inventory reduction, production costs, employee productivity, and asset optimization.
  • Role of the AI Team in Defining Success Metrics: Organizations where the AI team is actively involved in defining success metrics are more likely to use AI strategically. This involvement should be inclusive, encompassing feedback from groups managing data, business analysts, domain experts, risk management leaders, data scientists, IT leaders, and developers.

Gartner research separately shows that organizations where the AI team is involved in defining success metrics are 50% more likely to use AI strategically than organizations where the team is not involved. When selecting metrics, the AI team should include feedback from the groups that manage data, business analysts, domain experts, risk management leaders, data scientists, and IT leaders and developers.

  2. AI Value: Remove barriers to capturing AI’s value effectively

New tools like ChatGPT have turbocharged interest in the potential of AI, but to capture their value, executives need to look more broadly at business value, risk, talent and investment priorities and prepare for the potential disruption to existing business models and strategies. 

To date, AI business value has largely been generated from one-off solutions. Getting more value at scale, including from GenAI initiatives, may require deep business process changes; new skill sets, roles and organizational structures; and new ways of working. Failing to change will likely reduce your ability to capture the opportunities you identify.

Generative AI spells disruption to people, skills and processes

Map out how your organization will transform processes and systems and upskill people as GenAI becomes integrated into daily work. Deploying AI in a mindful and future-focused way will be the difference between long-term success and potential disaster.

Gartner strategic assumptions say:

  • By 2026, over 100 million people will engage robo colleagues (synthetic virtual colleagues) to contribute to enterprise work.

  • By 2033, AI solutions, introduced to augment or autonomously deliver tasks, activities or jobs, will result in over half a billion net new human jobs.

Identify issues that could slow adoption of GenAI projects or impede your ability to capture their value. Map out solutions and actions and assign an executive owner to champion the organizational change required. For example, if your organization lacks the data literacy needed to drive AI projects, incorporate executives (not just employees) into data literacy training and exercises, Make the chief data and analytics officer (CDAO) responsible for driving the program and ensuring other executives attend.

AI Risks: Prepare to assess and mitigate a range of AI risks

Government regulations and frameworks around AI are starting to emerge, so be aware of specific regulations in relevant jurisdictions. As AI usage continues to trigger questions about ethics and responsibility, new regulation may come in response to shifting public sentiments about AI use. In general, though, prepare for major types of risks, including:

  1. Regulatory. AI poses legal risks by potentially opening up organizations to lawsuits over copyrighted or protected content, information and data. Regulations are changing quickly, so be aware of local and jurisdiction AI regulations to ensure you stay compliant with governing policy. Also watch for industry-specific regulations, such as in life sciences and financial services.

  2. Reputational. AI can amplify bias and create a “black box” — an AI system with no user visibility into inputs and operations. Vendors that do not provide transparency on training datasets risk harmful outputs. Untested AI services can also pose risks through poor decision making and/or execution of tasks. Organizations need to build robust guardrails to prevent loss of intellectual property or customer data when building or buying generative AI services.

  3. Competencies. AI requires a unique set of skills that must be intentionally sourced through upskilling existing talent or from academia or startups. Skills in areas such as prompt engineering and responsible AI will be in growing demand in the near term.

AI threats and compromises (malicious or benign) are continuous and constantly evolving, so set principles and policies for AI governance, trustworthiness, fairness, reliability, robustness, efficacy and privacy. Organizations that don’t are much more likely to experience negative AI outcomes and breaches. Models won’t perform as intended, and there will be security and privacy failures, financial and reputational loss, and harm to individuals.

The Gartner AI TRiSM (trust, risk and security management) framework includes solutions, techniques and processes for model interpretability and explainability, privacy, model operations and adversarial attack resistance for its customers and the enterprise. We advocate standing up a cross-functional, dedicated team or task force, including legal, compliance, security, IT and data analytics teams and business reps, to gain the best results from every AI initiative.

Risks associated with generative AI

When it generates new versions of content, strategies, designs and methods by learning from large repositories of original source content, generative AI can result in:

  • False outputs. Generative AI can be unstable and erroneous in reasoning and fact, fail to fully comprehend context, have limited explainability and trackability, and be biased.

  • Security. Currently, any confidential information entered into public applications is stored and can be used to train new versions of the model. Sensitive data and intellectual property can become available to users outside the organization, including malicious actors.

  • Legal. Generative AI can present legal risks associated with intellectual property and privacy concerns, including copyright infringement, trade secret misappropriation, data privacy, model bias and model security.

AI Adoption: Prioritize AI use cases based on business impact and feasibility

In selecting use cases for AI, including those employing GenAI, line-of-business stakeholders should be able to clearly articulate the tangible business benefits they expect by asking:

  • What problem is the business trying to tackle?

  • Who is the primary consumer of the technology?

  • What business process will host that AI technique?

  • Which of the subject matter experts from the lines of business can guide the development of the solution?

  • How will the impact of implementing the technology be measured?

  • How will the value of the technology be monitored and maintained? And by whom?

Experiment first

Engaging in a comprehensive AI strategy without first experimenting with its component techniques puts the cart before the horse. 

Follow these five steps to introduce AI techniques:

  1. Use cases: Build a portfolio of impactful, measurable and quickly solvable use cases.

  2. Skills: Assemble a set of talents pertinent to the use cases.

  3. Data: Gather the appropriate data relevant to the selected use cases.

  4. Technology: Select the AI techniques linked to the use cases, skills and data.

  5. Organization: Structure the expertise and accumulated AI know-how.

This five-step formula is a tactical approach to the introduction of AI techniques, favoring a quick time-to-value perspective. It’s not a strategic, longer-term outlook.

Feasibility is as or more important than business value in use cases

Step 1 — identifying the most valuable use cases — should target concrete improvement projects coupled with tangible business outcomes. Feasibility is critical. 

Typically, returns are higher when risk is high and feasibility is low, but projects that are impossible to accomplish with available technologies and data aren’t worth pursuing regardless of the apparent business value.

Source- Gartner

Feasibility criteria include:

  • Technical. How well can the existing technology options improve the stated business use case to the level of “state of the art”?

  • Internal. Considerations such as (lack of) culture, leadership, buy-in, skills and ethics.

  • External. Considerations such as (lack of) regulations, social acceptance and external infrastructure.

A use case with an outstanding contribution to business value and easy feasibility is either a breakthrough or the market is missing a great opportunity. 

Data strategy affects feasibility of your AI projects

AI is very data-intensive, and while you can employ GenAI without integrating applications into your data stack, you won’t get the most out of AI without an enabling data strategy. 

Articulating clear data management and governance requirements, such as expectations for data quality and trust, lowers cost of data acquisition and helps you find and capture the data you need to power your AI. 

Also see: “Key Success Factors in Any Data and Analytics Strategy,” “Modernize Data Management to Increase Value and Reduce Costs” and “Becoming a Data-Driven Organization.”