Avoiding AI Missteps: Strategic Guidance for Executives
According to Mark Sims, Outthinker Networks Fellow and Managing Principal at Morningside Group, generative AI promises transformation, but companies must navigate implementation challenges and deepen their understanding to unlock its true potential.
As artificial intelligence (AI) promises to transform business, companies are racing to accelerate their AI investment and adoption. However, based on recent reports the hype has far from materialized as expected by investors. Key reasons are that many organizations are rushing to implement generative AI solutions without fully understanding the technology’s implications or potential. This has been a common path for new technologies – initial hype, followed by early frustrations due to outcomes not meeting expectations, with real outcomes coming much later in different ways than predicted.
We spoke to Mark Sims, an Outthinker Networks Fellow, former SVP of Strategy and M&A at The Scotts Miracle-Gro Company, and now Managing Principal at the boutique PE advisory firm, Morningside Group, to uncover what most companies get wrong when it comes to generative AI and what they need to do to get it right. He shared valuable insights into key considerations for companies and chief strategy officers (CSOs) for AI adoption.
“Companies need to ask: Do you have the data? And if you do, is there an actual high value use case?"
For chief strategy officers considering AI implementation, Sims recommends focusing on these key areas:
- Assess process understanding and data readiness
- Start with decentralized experimentation
- Build a business case
- Focus on applications
Assess Process Understanding and Data Readiness
Senior leaders making strategic AI decisions often don’t have visibility into the processes and sub-processes into which generative AI will integrate. Sims advises, “For each major process and sub-process, you need to know, what are the business rules that drive the activities around each of the nodes?” Before implementing generative AI solutions, companies need a clear understanding of their processes. Without this clarity, it’s challenging to effectively train generative AI models or agents.
New solutions are emerging to address this challenge. Sims notes, “Some experts in AI agents have ways to do that. You can have a camera that records an employee doing an activity, and the AI learns from watching, not necessarily learning from structured or unstructured data.”
However, data readiness remains essential. CSOs should assess whether their organization has sufficient, high-quality data to train AI models effectively. Sims emphasizes, “Companies need to ask: Do you have the data? And if you do, is there an actual high value use case?”
Insight
Before implementing generative AI solutions, companies need a clear understanding of their processes. Without this clarity, it's challenging to effectively train generative AI models or agents.
Start with Decentralized Experimentation
There can often be an organizational disconnect between C-suite enthusiasm and ground-level implementation. Sims observes, “I hear a lot of CEOs saying their companies are doing interesting things with AI. But people in middle management felt significantly less like they were doing something truly innovative with AI.” CSOs can play an important role in bridging this gap and ensuring effective AI adoption across the organization.
Despite the challenges, Sims strongly advocates for companies to start now with a decentralized, test-and-learn approach. “It’s better if you start putting the tools in the hands of businesspeople and have them figure out where and when they have a use case,” he suggests. This approach can help identify practical applications and build organizational competence. He advises companies to install the appropriate governance to protect proprietary and confidential information.
Sims adds, “There is no magic timeline, and the solutions will keep evolving. People should start using the tools. It is just a tool, like a spreadsheet or word document or Visio flowchart.” Once an interesting use case is identified, CSOs can help analyze whether there is a business case.
“There is no magic timeline, and the solutions will keep evolving. People should start using the tools. It is just a tool, like a spreadsheet or word document or Visio flowchart.”
Build a Business Case
Running generative AI models can be expensive, and companies need to consider the full cost of implementing and running these solutions. In many cases, the value equation hasn’t been solved yet. Sims cautions, “Do you have a situation where you’re replacing seven employees with an AI model that’s more expensive to run on an annual basis than those seven salaries?”
Organizations must ensure there’s a clear business case for AI implementation and verify the economic savings that stems from investing and running an AI solution. Sims explains further, “Green lighting a generative AI solution without a business case may seem like a way to force transformation or innovation in the business, but the upside economics of the initiative will eventually get it killed and the bad experience will likely create resistance for future AI initiatives.”
Insight
Organizations must ensure there's a clear business case for AI implementation and verify the economic savings that stems from investing and running an AI solution.
Focus on Applications
While generative AI is a powerful tool, Sims emphasizes that it’s the applications built on top of these models that will truly differentiate companies. “What will matter is not that one model is slightly faster than another. It’s what people build on top of these models. What will be the industry or domain-specific apps that people build on top of the LLMs?” CSOs should look beyond generic AI tools to industry-specific applications that can provide a competitive edge.
“What will matter is not that one model is slightly faster than another. It’s what people build on top of these models.”
Insight
CSOs should look beyond generic AI tools to industry-specific applications that can provide a competitive edge.
The Path Forward
While the potential of generative AI is immense, its successful implementation requires careful consideration and an experimental approach. CSOs must balance the need for caution with the imperative to start experimenting now. By focusing on these key considerations and fostering a culture of experimentation, CSOs can help their organizations leverage AI more effectively.
The goal should be to integrate generative AI not as a standalone solution, but as a powerful tool to enhance business processes, improve decision-making, and drive innovation. As generative AI continues to advance, the most successful organizations will be those that can adapt quickly, learn continuously, and strategically integrate AI into their businesses through real business cases.
For CSOs, the challenge – and the opportunity – lies in guiding their organizations through this transformative journey, ensuring that AI investments are strategically aligned, economically viable, and effectively implemented across the organization.
About Outthinker Networks
Outthinker Networks brings together two executive peer networks – the Outthinker Strategy Network and the Outthinker Innovators Network – to help senior strategy and innovation leaders solve their most pressing challenges and keep their organizations ahead of the pace of disruption.
Are you a strategy, innovation, or transformation executive interested in joining one of our networks?
Authors
