Terryel Hu, Ph.D.

I was recently at a board meeting where a group of executives from a small financial services company gathered to discuss AI. This wasn’t just an idea - they were committed to making investments in something tangible. Like many, they were drawn to the concept after noticing announcements made by competitors. They’d made a remarkable move by staying in tune with the market. They understood the changing landscape and gathered their most experienced directors for this meeting.

The meeting was productive, to say the least. Each division leader was tasked with creating a roadmap. Towards the end of the meeting, each leader was asked about their approach. They were able to walk through their plans in incredible detail. They congratulated one another, headed back to their teams, and began creating the plan.

When “Yes” Does Not Mean “Yes, We Can Do It”

Then it struck me. The executives, though brilliant, were creating a plan in isolation. They could provide the roadmap to product–market hypotheses, experimentation, prototyping, and launch. The members at that meeting were clearly more innovative than most. However, they developed a plan isolated from other parts of the business. There was little input from IT, marketing, or the people servicing their products.

I asked one of the division leaders what his team thought. He said they were thrilled with the initiative. That success, however, was short-lived. The technical teams realized the company lacked the infrastructure to execute the “AI initiative.” They also encountered errors that were unrecognizable during the trial phase.

So their plan eventually hit a roadblock. That’s normal. The surprising part was that they planned in isolation. The more brilliant the plan appeared, the less they sought validation from others.

Here's what's happening: The more consensus an idea gets, the more one feels confident planning it alone. It can be misinterpreted as a sign to continue. But a “yes” in such cases is only a half-truth. A “yes” from an IT department does not mean “yes, we can do it.” Similarly, when a marketing team says, “Yes, AI is a game-changer!” it doesn’t mean the company can actually deliver it.

Planning is a Team Sport

Adding the word “strategic” in front of planning has perhaps made it worse. People have come to take “strategic planning” to mean “a plan that’s limited to the strategy table.” That might work in some instances, but it rarely proves successful in large companies where the product depends on several departments. Too often, plans are formulated at the top, and it’s expected that the rest of the organization will implement them.

Even though this top-down approach never goes as smoothly as planned, it remains the primary method companies use to encourage employees to adopt technology. Here’s what it looks like: A new initiative is created by top management teams. Employees are told that the new way of working will bring opportunities. If they show resistance, it must be overcome through a mix of frameworks and surveys. If the technology isn’t adopted, the company reframes it as a compliance issue.

How does this top-down approach lead to isolated planning? Well, if you look at those activities, you’ll notice that the process can be carried out by a single person. That’s what I’ve repeatedly seen. That one person carries out the plan with minimal contact from others. They accept the “yes” and ignore those who say “no”. They continue down this track due to fear of uncertainty or, at worst, the fear of being wrong. In doing so, they avoid valuable feedback and miss out on the opportunities to improve the success of the plan. Those working in competitive industries know this well - they view every piece of feedback as a chance to outshine competitors.

If we want initiatives to succeed, we must treat strategic planning not as a solo project, but as a team sport. After all, no one wins alone. We often call it “strategic planning” because it’s formulated at the boardroom table, but that doesn’t prevent the planning process from being collaborative and problem-driven. The planning, no matter how brilliant, will fall short if developed in isolation. To succeed, we must consider the contributions of other teams - whether it’s IT, marketing, operations, or frontline employees - and what roles they play.

The heavy-handed approach would be to block governments and civilians from DeepSeek. The United States, South Korea, Australia and Italy have imposed some form of ban on DeepSeek. China already has a censorship policy which blocks Google, YouTube, Vimeo, Facebook, Yahoo, Wikipedia, Reddit, Netflix, Spotify, Steam, Linkedin, X (Twitter) and Skype. Major news media sites are also blocked.

Western countries are not banning it due to racism, nor a full-blown arms race. The reality is that the adoption of AI depends on how well we can regulate it. It involves understanding the benefits and risks. Does data privacy, storage and handling align with the security standards of our nation? What is its social and economic impact? Do the benefits outweigh the costs? These are questions that policymakers normally ask. However, the capabilities of AI have outpaced our ability to address these questions. There are too many unknowns, and that is where the AI hype meets reality.

One of the barriers that will stop DeepSeek, or any foreign technology, will be the security risks that it imposes. It so happens that DeepSeek is from China, a global superpower, that is in direct competition with the US. But if we look at the other AI competitors, it’s only a matter of time before other countries release their version of Open AI. It won’t be a race to limit each other’s progress, but a race to create “rules of the game”. They need proper AI policies and frameworks to compete.

Beyond the Plan: Why Strategy Is a Continuous Learning Process

Strategy is often seen as a static roadmap, but in reality, it is a continuous learning process. The best organizations don’t expect their plans to stay fixed. They move forward by learning and discovering new ways to improve business performance. In practice, no one implements a plan without expecting it to change. Although many agile practices have become faddish, it has one principle that remains relevant today: It’s naïve to expect the world to stay still while you write up a plan.

Having a fixed plan is wishful thinking. Top innovators will know that, by the time they finish writing a plan, the variables will have changed. To use a sports analogy, strategy can be considered the gameplay. Like any competitive sport, that gameplay evolves as the team gain new insights, tests ideas, and adapts to unexpected moves from the opposition.

The Myth of Strategy as a Fixed Plan

Traditional strategy still conjures images of boardroom meetings, long-term plans, and rigid frameworks. Companies that cling to outdated processes risk irrelevance in fast-moving industries and will see their favorite tools become obsolete. The classic red, yellow and green risk chart doesn’t seem to respond well to the changes in AI capability. How would they assess the impact of AI, if we are unsure of its capabilities? Should Agentic AI be medium or high risk? How will they assess the risk of LLMs when it never existed in their industry? In such cases, strategy should not be a fixed blueprint but instead a dynamic process of discovery and refinement.

A strategy around ‘discovery and refinement’ is still very intentional. For example, some companies launch pilot programs or prototypes to test market responses before committing to full-scale implementation. Others prefer iterative processes, where small teams experiment with different solutions, gather user feedback, and adjust their products. Maintaining this cycle of discovery and refinement is also called organizational learning.

The Learning Organization

Senge’s learning organization emphasizes five traits that companies should cultivate and where continuous learning is embedded into the culture.

• Personal Mastery. While mastery often relates to the 10,000-hour rule, this estimate doesn’t do much to motivate learners. What motivates them to build mastery is the opportunity to experiment. Businesses that prioritize experimentation and feedback tend to develop new products and services that align more closely with evolving customer expectations. To improve mastery, employees require an environment where they can test ideas, refine approaches, and implement new insights.
• Mental models. Our way of thinking is ingrained in assumptions about the world. They are so taken for granted that simple reflection is not enough. The assumptions behind our reasoning must be challenged. Leaders must shake the habit of correcting errors with “yes or no”, and engage in evaluating the assumptions driving the error. Suppose we had a product defect. Would it need to cause that product defect? What design, policy, or management decisions led to that defect?
• Shared Vision. Capabilities may be different, but what brings a group together is the vision they share. Unfortunately, this is also why cross-collaborations fall short of expectations. Employees from IT and employees from marketing strive to hold a different vision of success. For that, we must avoid looking that stereotypes and focus on how they have evolved. We must be curious and ask questions like “Where have IT and marketing teams succeeded together?” A close look at the tech industry will see a fusion of the two — product managers, adoption specialists and technology marketing specialists. Focus on the practices, cultures and activities that enable cross-collaborations.
• Team Learning. Companies that encourage diverse teams to work together generate richer insights and innovative solutions. Bringing together different perspectives enhances problem-solving and strategy development. However, this doesn’t happen by putting random employees into a room and asking them to learn from each other. Employees must feel safe before they share insights, challenge assumptions, and take risks without fear of punishment.
• Systems Thinking. It can be easy to get lost in a large company. It’s built on systems that don’t ‘speak to each other’. Let’s say a company has acquired two distinct software programs: a customer tool and an employee tool. They may not produce the same data, but they are interrelated. Workers can still leverage insights from both tools, linking employee performance and customer sales. The messiness of a system can provide leaders with a chance to form new connections, insight and strategies.

Effective leaders understand that strategy is not a one-time event but a continuous journey of learning and adaptation. The smartest strategy is the one that keeps evolving. As business environments grow increasingly complex, the ability to learn faster than competitors will define the winners.