Tailored business solutions. Organizations can implement generative AI tools to create customized solutions that align with strategic goals.
There are distinct types of generative AI implementation, according to recent research from MIT-CISR led by Nick van der Meulen and Barbara H. Wixom. The study delves into the essentials of effective management for the delivery of marketing and business value through generative AI.
The first type is broadly applicable generative AI tools that individuals use for multiple purposes, and the second is specialized generative AI solutions tailored to solve specific organizational purposes.
The latter solutions are designed to align with strategic business objectives. To do this, they must be integrated seamlessly with existing processes, systems and offerings -- their aim being to enhance marketing and business effectiveness and drive meaningful outcomes.
Generative AI tools change how users interact with technology. These multipurpose tools lack predefined use cases. Instead, they evolve through user engagement, allowing users to enhance their personal productivity. Whether drafting emails, brainstorming ideas or analyzing data, generative AI tools support users in streamlining workplace tasks and increasing efficiency.
Successfully implementing generative AI tools depends on many factors, including employee education. Wixom and van der Meulen wrote that users need to learn how to give clear, specific instructions -- including clearly defined problem statements, context-specific objectives and detailed descriptions of desired processes. For example, a marketer trying to generate innovative campaign ideas should provide the tool with a comprehensive overview of the campaign topic, examples of past successes, target audience, desired tone and intended outcomes.
To achieve success with generative AI tools, van der Meulen and Wixom wrote, business and marketing leaders must establish clear usage guidelines and guardrails. These guidelines should outline permissible tools and their conditions for use while highlighting associated risks and potential consequences. A robust generative AI policy should focus on data input and output risks, clearly delineating "always okay" use cases and "never okay" use cases. "Never okay" use cases include the use of personally identifiable information, strategic insights or proprietary data.
For this reason, marketing and business leaders should foster an environment where employees can harness generative AI tools effectively. Establishing AI direction and evaluation practices empower employees to derive maximum benefits from these innovations.
Finally, it's a good idea to narrow down your choices to a few select vendors. Forming a cross-functional team of potential generative AI tools, users can assist IT in identifying the most promising options for the organization's needs.
Related Article: A Game Plan for Generative AI in Customer Experience and Marketing
Generative AI solutions are developed based on specific business goals, aiming to meet strategic objectives while providing clear benefits to key stakeholders. For instance, a generative AI solution designed for a call center could leverage large language models (LLMs) to analyze the content and tone of customer-agent conversations and offer real-time coaching to enhance agent performance.
This approach not only helps improve operational efficiencies but also increases revenue through heightened agent productivity and customer retention, illustrating the potential for substantial financial returns. As demonstrated by recent innovations like Salesforce's Agentforce, the scope for generative AI applications is expanding, creating opportunities for even greater value generation.
However, implementing generative AI solutions presents its own unique challenges. First, as employees at all organizational levels explore how generative AI can enhance processes, systems and offerings, the demand for new generative AI solutions grows. This can lead to a fragmented approach that complicates the overall integration of these tools.
Second, most foundational models that underpin generative AI solutions are owned by a handful of large vendors who control critical aspects such as model mechanics, distribution and usage rights. This concentration, wrote van der Meulen and Wixom, can create limitations for organizations seeking to leverage these technologies effectively.
Finally, the ultimate value derived from generative AI solutions is highly contingent on the chosen development approach, underscoring the importance of strategic planning and alignment.
To succeed with generative AI solutions, organizations must establish a formal, transparent innovation process that incorporates clear governance structures and encourages early engagement with stakeholders. This collaborative approach ensures that the solutions developed are not only scalable but also aligned with the organization's strategic objectives. By prioritizing rapid prototyping and strategic alignment, companies can effectively vet their solutions while minimizing risks associated with shadow generative AI efforts, ultimately maximizing value across the enterprise.
Moreover, organizations should formulate comprehensive guidelines for generative AI development decisions and create a robust vendor partnership strategy. By outlining clear criteria for selecting generative AI partners and establishing frameworks for collaboration, organizations can navigate the complexities of the generative AI landscape more effectively. This structured approach not only facilitates the successful implementation of generative AI solutions but also allows organizations to leverage their full potential for innovation and growth.
Generative AI offers leaders two key opportunities -- first, to improve individual productivity by equipping employees with generative AI tools, and second, to drive business growth by implementing tailored AI solutions that transform processes, systems and products on a large scale. Leaders can cultivate a virtuous cycle in which heightened employee knowledge and skill with generative AI tools inspire more innovative generative AI solutions.
The best way for organizations to begin is by selecting a few generative AI tools from trusted vendors. This not only facilitates experimentation but also lays a solid foundation for realizing future value. By carefully choosing which tools to implement, organizations can streamline their efforts and make sure they are effectively harnessing generative AI's potential.
Organizations can consider various approaches to implementation, and each approach offers distinct advantages. The "buy" approach involves vendors providing, running and maintaining the generative AI solution, allowing organizations to adopt the technology quickly without needing to invest in model development or fine-tuning. While this approach offers speed and convenience, off-the-shelf solutions may be opaque and tailored to narrow contexts, such as specific functions or industries.
A "boost" approach occurs when the vendor retains responsibility for the model while the organization enhances it using proprietary data. This may involve fine-tuning the model for improved performance in specific contexts or using retrieval augmented generation (RAG) to add relevant contextual data to user prompts. While RAG adds context to otherwise opaque models, it also increases prompt length and usage costs, necessitating collaboration among data scientists and domain experts to refine the model's output.
Finally, a "build" approach means that the organization fully assumes responsibility for developing, running and maintaining its generative AI solution. While this option can lower ongoing usage costs, it demands significant upfront investment in computational resources and advanced data monetization capabilities, including effective data management, data science expertise and robust AI explanation practices.
Ultimately, the chosen approach should align with an organization's strategic objectives, capabilities and resources to ensure the successful integration of generative AI into the corporate business model.
Related Article: How Generative AI in Customer Experience Is Reshaping the Landscape
Generative AI offers leaders two key opportunities for business success: improving individual productivity with AI tools and creating customized solutions that transform processes and drive large-scale financial returns. By strategically selecting a few trusted generative AI tools, organizations can empower their workforce and encourage a positive cycle of skill growth and innovation.
Implementation approaches vary. Each approach, however, has unique benefits and challenges, requiring organizations to align their strategy with available resources and capabilities to maximize the value generated for marketing or corporate initiatives.