As businesses begin exploring generative AI, a common question is “where do I start?” Consulting company McKinsey provided some recommendations in their 2023 report “The economic potential of generative AI: The next productivity frontier“, including what roles may benefit the most from generative AI – and the answers may surprise you.
Where to start?
For their report, McKinsey analyzed “63 generative AI use cases spanning 16 business functions”, then considered how those use cases could affect over 850 different occupations. They estimated that generative AI could add “the range of $2.6 trillion to $4.4 trillion in economic benefits annually when applied across industries”. This is on top of the $11 trillion to $17.7 trillion added by *non-generative* AI, such as such as advanced analytics or machine-learning systems.
From this, they determined that 75% percent of the value from Generative AI would come from just 6 different use cases (from highest to lowest value):
- Sales
- Software Engineering (for Corporate IT)
- Marketing
- Software Engineering (for product development)
- Customer operations
- Product R&D
Reaching Customers More Effectively with Generative AI
A major advantage of generative AI for marketing and sales is the ability to quickly and effectively tailor messages to specific audiences – and even specific customers, if you have enough information. GenAI also has a powerful place in lead development, with the potential for quick and early lead qualification before a salesperson gets involved.
Additionally, generative AI can serve as a powerful creativity assistant for marketing activities, including;
- Analyzing complex, unstructured data for trends and insights
- Helping marketers consider a greater range of concepts and strategies
- Creating mock-ups quickly
- Using AI personas for virtual customer round-tables
- Fine-tuning messaging to be more succinct, persuasive, and discoverable
Product Development with Generative AI
The benefits of generative AI for product development come in two main areas.
First, as a discovery tool. The same sort of process that undergirds large language models – pattern-matching based on detecting higher-level structure – can be applied to other systems that a) can be represented symbolically and b) have well-defined rules for relating symbols. For instance, many companies are developing GenAI-based systems for chemistry and life sciences products, providing a more creative and flexible alternative to existing computer-based models.
Second, generative AI can also be used as an assistant to product designers to help accelerate their work. The report calls out the potential for more efficient material choices, better design for manufacturing, development of documentation and test plans, and perhaps even improving product appeal. Product designers will still be critical to product development, but GenAI can reduce some of burdensome overhead of the role, allowing designers to focus more on the strategic aspects of the product.
This development assistance applies to software, as well, where GenAI has proven itself as a powerful tool for software development. Given that GenAI can both write code and analyze error logs, some companies have even started developing “fully-autonomous” software agents, like Cognition AI’s Devin, that can theoretically create a whole software system based on some basic high-level guidance. These tools, including Devin, still don’t fully live up to their promise yet, but even their limited effectiveness today show the potential in the near future.
Operations and Infrastructure with Generative AI
Customer service is a major target for acceleration with generative AI. The report cites research showing that one company with 5000 customer service agents used GenAI to increase issue resolution 14% per hour and reduce handling time by 9%, along with reducing agent attrition and issue escalation by 25%. For customer self-service, the report notes that automated systems already handle approximately 50% of customer inquiries, and GenAI could cut the remainder in half again.
A critical note here: you must be very careful when allowing your customers to interact directly with the AI. Even commercial LLM’s don’t always answer questions correctly, and can cause embarrassing incidents that your company will still be accountable for. Given today’s technology, a far safer bet is to empower your customer service reps with GenAI tools, letting them manage accuracy and the relationship while the AI manages the data.
GenAI can also help with accelerating IT support tasks through their coding capabilities. Many DevOps, cybersecurity, and other support tasks rely heavily on software scripts that coordinate their tasks. GenAI can both assist with developing those files, as well as analyzing log files for issues and recommending improvements for key internal processes.
The Other 25%
The fact that the above roles cover 75% of the value add doesn’t mean other roles in the organization will remain the same. In fact, McKinsey highlighted that the majority of roles could see some benefit from generative AI.
So where should you start as you explore expanding roles with Generative AI? Start with the 75%, and expand from there.
Want to stay informed on the most important AI news in just 30 minutes a month? Sign up for the Executive Summary newsletter by AI For Business.
If you liked this post, have future posts delivered straight to your inbox by subscribing to the AI For Business newsletter: https://buttondown.email/AI_For_Business. Thank you!