Cost-Benefit Analysis: Is Generative AI Worth the Investment for Enterprises?

At MachineCon 2024, hosted by AIM Research, industry leaders gathered to dissect a pressing question facing businesses today: Is the investment in generative AI worth it for enterprises? The panel discussion brought together experts from diverse sectors to share insights on cost-benefit analyses, implementation strategies, and real-world experiences with this transformative technology. The panel was moderated by Gabi Steele, Co-Founder & CEO at Preql along with panelists Deepak Jose, Global Head and Senior Director for One Demand Data and Analytics Solutions at Mars Wrigley, Zee Moradi, Chief Data Officer at UBS Bank USA, Bhaskar Kalita, Global Head of Financial Services & Insurance at Quantiphi and Dwarika Patro, Founder & COO at Aays.

Starting Small for Greater Impact

A key takeaway from the discussion was the importance of beginning with targeted, high-value use cases rather than attempting an extensive, company-wide AI overhaul. Experts emphasized that focusing on a few “hero products” or programs where AI can deliver significant benefits allows organizations to demonstrate tangible results and build momentum for broader adoption. For example, a major consumer goods company ran an ideation session that generated hundreds of AI-related ideas, ultimately narrowing them down to 50 high-value initiatives. This methodical approach ensures long-term success and sustainable adoption.

The Three Pillars of AI’s Economic Impact

Generative AI’s economic impact can be categorized into three main areas: revenue generation, efficiency gains, and productivity improvements. Although revenue impacts are still largely speculative, tangible gains in efficiency and productivity are already evident. For instance, a services organization reported a 30% increase in productivity, particularly in tasks such as code migration and generation. A data transformation project that would traditionally take 15-16 months was completed in just 9 months using generative AI, highlighting significant time and cost savings.

Addressing Human Factors in AI Adoption

Despite technological advancements, challenges often arise from human factors. For example, an insurance industry case highlighted a scenario where a new AI tool for claims processing faced resistance from experienced adjusters despite high accuracy scores. This underscores the importance of change management and considering the human element in AI adoption strategies.

Managing ROI Expectations for AI Investments

A long-term perspective is crucial when evaluating AI investments. While some benefits, like those in code migration, can be realized quickly, more complex projects may take 18-24 months to show significant ROI. Industries such as consumer packaged goods (CPG) often experience transformational changes on a two to three-year cycle. Thus, having a clear, long-term vision while demonstrating incremental wins is essential.

Navigating Infrastructure and Operational Costs

Infrastructure and operational costs are pivotal in the cost-benefit equation. However, these costs are expected to decrease over time as technology improves. Experts advised incorporating anticipated cost reductions into business cases and ROI calculations. They noted that current pricing models for large language models, often based on tokens, present challenges in translating these costs into real-world applications. As technology evolves, similar to the cloud computing industry, clarity on costs is expected to improve.

Sustainability Considerations in AI

Sustainability was a dual focus in the discussion. Companies are leveraging AI to achieve better sustainable outcomes, such as enhancing crop yields and developing eco-friendly packaging. For example, one food company is using AI for sustainable cocoa farming and supply chain visibility. Additionally, responsible AI use itself is becoming a priority, with organizations partnering with cloud providers committed to carbon-neutral data centers. Aligning AI initiatives with broader corporate sustainability goals is crucial.

In highly regulated industries like financial services, managing risks around model performance is essential. Without robust metrics and governance frameworks, AI investments could be ineffective or even detrimental. Ongoing monitoring and risk management are critical to successful AI integration in these sectors.

Strategic Frameworks for Aligning AI with Business Value

Aligning AI initiatives with core business problems and value creation is crucial. Experts recommended using frameworks to categorize AI projects based on ease of execution and potential value. One approach involves a 3×3 matrix that classifies initiatives into “do things,” “do things better,” and “do better things,” helping organizations prioritize investments effectively.

The Transformative Potential of Generative AI

Generative AI holds the potential to bridge gaps between business functions and technical teams, potentially revolutionizing the software development lifecycle. Natural language processing and code translation were highlighted as particularly promising areas. AI is already being used to translate complex technical requirements into code, potentially transforming the software development process.

Caution Against Hasty Implementation

Despite the excitement surrounding AI, it’s crucial to avoid rushing into implementation due to market pressures or competitive fears. A strategic, thoughtful approach is necessary to maintain credibility and avoid setbacks. Anything done hastily might undermine broader community confidence and set an organization back years.

In conclusion, while generative AI offers significant opportunities for enterprises, success hinges on careful planning, strategic implementation, and a willingness to start small and scale gradually. Organizations must balance the drive for rapid innovation with the need for responsible, sustainable AI adoption that aligns with their long-term goals and values.