Want to lead with AI? Drop your AI Strategy and focus on these four planks instead
October 2025
Executives tend to ask me two questions when it comes to AI: Are we already behind, and where are we compared to our peers?
Whether these are useful questions is irrelevant, as all are asking it (as are their Boards).
Rankings of where companies are in AI capability and adoption vary, but the lack of widespread adoption is dramatically underestimated and understood by tech evangelists. Organizations such as Gartner, produce a Magic Quadrant report within specific markets for AI rather than an overall analysis (such as conversational AI platforms). Press outlets such as AI magazine have their annual top 10 and Forbes ranks the top 50 private companies in AI. Researchers and universities weigh in with more robust methodologies: Stanford produces The AI Index 2025 Annual Report, IMD ranks the world’s leading companies on AI adoption and use, and Harvard’s Digital Data Design Institute spotlights recent research.
As with any rankings, these vary based on influence and motivation, and lacking insider information, there is a lot of guesswork. At a macro level, on average, technology companies are leading. Financial services are leading insurance, and manufacturing and retail lagging, with construction even further behind. The significant distribution within each industry is more of the story, though, as is the significant adoption variance inside individual companies.
Rather than sending you to-do lists, one way I can assess status and progress is by asking two questions back: Do you have an AI strategy, and if so, who owns it?
If an AI strategy exists, and lives under the CTO/CIO without broad awareness within the organization by business unit leaders, I have a telling data point. You have a plan for technology but likely lack a pathway to true transformation.
When we are trying to lead breakthrough growth strategy in a way that delivers impact, we must dramatically accelerate how we create value.
AI is a moving target. Definitions change as the technology evolves: that is part of the challenge and fun of teaching and advising executives on it.
Strategy and value creation are different.
The definition of growth strategy is as powerful as it is constant: strategy is an articulation of how an organization creates value, and value creation is about how an organization drives the biggest gap possible between two levers: customers’ willingness to pay for your products, services, and solutions and your total cost of delivering that value.
How we achieve value creation adjusts, the definition of strategy as value creation has not.
So what’s the definition of AI Strategy?
It doesn’t exist.
AI AND Strategy:
Now is the time to pause to reflect and revisit your existing strategy to ensure you are not just taking advantage of AI’s current opportunities but also strategically preparing for future advancements and mitigating potential threats.
This does not mean developing an AI strategy, as if it were a special case independent of the rest of your business. You don’t need an AI strategy: you need a value-creating strategy for your business that takes into account AI’s impact and your beliefs about where it will go in the future. There is no such thing as ‘AI Strategy.’ The role of AI is to enable, support, and accelerate value creation.
It's probably worth pausing for a critical strategy reminder. You set strategy at the level of your organization where you create value. The entity that needs a strategy is your organizational entity operating with limited resources in a competitive environment. There are two such entities: corporations, which need a corporate strategy to enable them to make parenting decisions, and business units, which need a strategy to dictate how that operating unit will create value in the strategic environment. The role of the different supporting and enabling parts of an organization is to make this strategy work, ideally in a differentiated way.
Punchline: If your ‘AI strategy’ lives separately from your value creating strategy you can have internal competition not convergence.
Leading and bringing AI (and its related digital technology components) into your value creation strategy is critical, and you can do so by excelling at four interrelated planks in an optimized way.
AI Approach: Four planks
Four main planks are critical for your strategic AI approach, and these should form part of your regular Board updates.
Technology and data Foundation: Tech & Data foundation (infrastructure, AI factory, data lake, systems, etc.)
Capability building: People skills and capabilities and organization capacity (such as processes) to lead and transform with AI
Governance and risk management: Cyber, risk management, and governance principles
Value creation Outcomes: How you will link AI/GenAI/agents to enable, support, and accelerate value creation
Let’s look at each in turn.
Tech and data foundation: Underlying any successful investment, and eventual transformation, with AI necessitates a solid foundation of technology and data. Even smaller organizations often have competing data lakes, antiquated systems, and legacy tech debt that begins slowing down any future improvements. This plank is about understanding your data and its link to operational and competitive advantage and building a pathway to being able to leverage it, reconciling systems and processes, and ensuring you have a scalable technological infrastructure in place.
-> Your role as a leader is understanding the current status, defining critical bottlenecks, and then build an aligned pathway to progress.
Governance and risk management: AI transformation needs risk management and governance in place, which is often referred to as RAI (responsible AI). A paradox of governance is that clear boundaries actually provide freedom. Lacking a clear, coherent, and usable RAI framework stifles employee creativity and constrains progress. Employees know there are guardrails in place, and your talented team members avoid possible landmines in execution. This means they do less experimentation, innovation, and learning as they are unsure of where the boundaries are, or they utilize their innovative capabilities, but outside of work. What is also risky for making progress with AI is an RAI approach which does not take into account the current reality of what the tools can do. An example would be limiting all public LLM use without providing custom GPTs or restraining the entire workforce to one tool, such as Microsoft Copilot, without differentiating between tasks uses and teams, such as coding versus content generation.
è Your role as a leader is to define what RAI means for your organization, putting critical safeguards in place (such as protecting against sensitive data leaks or shadow IT), and ensuring your cyber security and related governance components are fit for purpose while allowing enough freedom for employees to move.
Capability building: Just as other skills are needed to excel, building critical capabilities in AI is a new language across your organization. To not provide capability building and expect employees to learn it on the job will lead to slowed progress and frustration. You likely have self-motivated team members who are racing away, training themselves on the tools and integrating them into work. You also likely have resisters who cannot yet see the value or the need to upskill on something new.
->Your role as a leader is to commit your time, treasure, and talent to upskilling the organization in a way to continues to grow those who are already ahead without leaving behind critical employees you want to bring on the journey. This likely involves working with a partner or set of partners, but this will take investment versus letting people figure it out themselves or expecting it to happen as part of their day job.
AI AND Value creation:
As above, the role of AI is to enable, support, and accelerate the organization’s value creation strategy. The three planks above help you do this at scale.
Enabling forces are the foundation to having success with AI (data analysis, accessibility, productivity). Supporting forces are directly enhancing each of the strategic priorities (that is, taking each top priority or must-win battle and considering how AI can support its success. Accelerating forces are those that dramatically change the steepness and velocity of the growth and performance curve of the organization.
Where to start?
A few ways exist to identify where to integrate AI into strategy: by strategic priority, by your value chain, by individual team or role needs, or by line item in the P&L.
All work to a varying degree, but I caution against the most popular: role based. Letting individual teams select their own tools and use cases that improve their personal productivity at first appears to be working. It boosts individual employee morale and often productivity. The challenge is that one person’s gains is often another team members’ lack of productivity (see the recent HBR on workslop here). The bigger issue is it leads to duplication of tools, lack of coherence across teams and units, wasted resources, and what I call ‘science projects’ or fun initiatives spreading across the organization that do little to affect change.
It may be an ok place to start, but as a leader, you need to move your organization from personal productivity to process reimagination and transformation.
To bring AI into strategy, I prefer to start with the top priorities, or must-win battles, and then do a vale chain mapping.
First, go through each of your must-win battles to identify the biggest opportunities to leverage AI; this is much preferred rather than having a separate MWB on ‘AI’ (this will lead to a solution chasing problems).
Then, map your value chain. As a recap, your value chain is the series of interlinked activities that allow you to create value, or every major chunk of activities from raw ideas and raw parts to happy customers. This includes the primary functions (such as R&D, supply chain, production, customer experience) as well as enabling functions (IT, Finance, Legal, HR).
Bring together a small, diverse team (not just from IT/Data Science but also not excluding them!) and walk through each step of your value chain to identify major friction points (where things are not working as they are or value is being left on the table) and other quick wins. Start with opportunities and not use cases, as you can narrow your thinking too much and too quickly by starting with what AI can do; instead, start with value creation opportunities.
After you’ve identified the opportunities, link these to use cases, where you may need to rely on the knowledge of folks with more technical backgrounds (or partners). Here you want to identify how these opportunities can be solved with AI and how you will measure impact.
Then bucket all these opportunities into ‘needle up’, or those that dramatically increase customer/consumer/client willingness to pay and ‘needle down’, or those that dramatically reduce cost or increase efficiency. Why this is needed is it is too easy, especially in the earlier days, to focus only on ‘needle down’ use cases, as these are easier to identify, measure, and match to existing AI applications. However, while productivity and efficiency and cost reduction are all important, in some ways gains that can be achieved with ‘needle down’ are finite whereas (theoretically) the gains from needle up opportunities are infinite, or at least not maxed.
Once you have these, prioritize by impact and effort as well as by fit with our existing technology (see the first plank above) and begin a pathway to execution.
Note: If you need a refresher on value chain mapping, read chapter 5 of the SRT book, and chapter 8 of the SRT book will remind you how to set your top priorities, or must-win battles.
Parallel pathing and creating new competitive advantage
The challenge with these planks is they are not sequential. The way to achieve success is by the uncomfortable execution task of parallel pathing across all four.
This will not feel intuitive at first. Where companies tend to stumble is when they prioritize one or two of these areas above the others. One of your team members will want all the governance in place before any capability building starts. Another will want capability achieved before deploying any tools. A third will argue nothing can be done until the tech stack and data pipeline is fixed. Each will make a valid argument, but this is the trap you must avoid. You must move even though it often feels premature.
This is what sets leading companies apart: they are executing across all four planks in parallel.
For those of you that have Advisory Boards, there is a related challenge in that most directors have areas of expertise in one of these planks, often at the expense of others. They will want to steer the conversation there, so you will need to control the dialogue. This is why I encourage CEOs to select words for each of these four planks and keep them consistent over time. Never underestimate the potential for confusion (from lower-level employees to senior leaders to Boards) and consistency of wordage is key.
Your AI pathway to advantage
Intersecting all these planks is the role of data. Most companies dramatically overestimate the power and value of their data; especially as it relates to their competitive advantage. Many more companies will dramatically underestimate how quickly AI is eroding their existing competitive advantage, a topic for one of my next blogs (but you can ask me for an early copy).
AI and the practice of using computational methods to mimic what was historically an innately human task of problem solving and driving efficiencies is not new to company planning. We have been using the term since the late 1950’s, and as it progressed from earlier iterations of programmed machines through advancements in machine learning and deep learning, organizations continued to advance and deploy it – to varying extents – within their digital initiatives. While recent advancements in generative AI have pushed this onto the executive agenda, for most organizations AI is still an operational tool – living in targeted use cases in distinct business units or a part of the prerogative of the Chief Digital or Technology officer. What many companies miss out is thinking through and acting on the broader implications of what it could mean for their business over the years to come. Unless they do so now, it may be too late.
A reset is needed with AI. Some organizations face up to turbulence, stabilize and survive. Others do this then carry out a reset and go on to thrive. What sets this group apart? They combine a strong balance sheet with strategic insights from the reset and then execute with agility and learning. Agility is making good decisions quickly that are aligned with strategy, and using AI tools allows teams to do this even better and faster. Learning velocity measures how quickly knowledge acquisition occurs, is shared, transferred, embedded, and used to further performance.
Things will continue to change quickly, much will be commoditized, but some organizations will set themselves apart with the differentiator of learning velocity. Those that learn fastest will win.
Learning starts by asking the right questions. In strategy, it starts with the fundamental question: Has the situation changed for our current strategy given AI’s recent advances? If, as is likely, the answer is ‘yes’, the next question is ‘is our leadership team willing to have the tough conversations about what needs to be reset, and are we prepared to build our foundation, governance, and capability to execute with more alignment, agility, and learning going forward?’
Much hangs on the answers and your commitment as a leader to regain AI (and GenAI, agentic workforces) as part of strategy, not alongside or in opposition to it.
