AI Roadmap Series: Step 1 - AI Strategy

AI has become the word in tech for the past year and a half. Every executive wants to implement something with AI. Budgets are often allocated, but what's missing is clarity on what to actually implement.

That's where a structured roadmap comes in. In this series, we'll follow Gartner's 7 steps for an AI Roadmap. These steps can guide you before you invest in any AI project:

1. AI Strategy
2. AI Value
3. AI Organization
4. AI People and Culture
5. AI Governance
6. AI Engineering
7. AI Data

Today, we begin with the foundation: AI Strategy.

Why Strategy Matters in AI

"Strategy" is one of those overused words—often misunderstood and sometimes poorly executed. But when it comes to AI, starting with a sound strategy is the difference between meaningful impact and wasted investment.

If you want to go deeper into strategy itself, I recommend Good Strategy, Bad Strategy by Richard Rumelt. In it, Rumelt introduces the "Kernel" of a good strategy:

1. Diagnosis: Define the nature of the challenge and identify the critical obstacles.
2. Guiding Policy: Establish an overall approach to overcome these obstacles.
3. Coherent Actions: Take coordinated steps, backed by resources, to implement the policy.

By contrast, bad strategies usually fall into one of these traps:

• Fluff: Buzzwords and slogans dressed up as strategy.
• Failure to Face the Problem: Ignoring the real obstacles.
• Mistaking Goals for Strategy: Confusing high-level ambitions ("be number one") with a plan to get there.
• Bad Strategic Objectives: Scattered goals without connection or focus.

The truth is, good strategies are often simple, focused, and leverage an organization's strengths against the most critical problems.

Translating Strategy into AI

When applied to AI, strategy requires clarity and discipline. As Seneca said, "The first step in the learning process is the unlearning of error." Most AI strategies today are nothing more than fluff, so avoiding that pitfall is crucial.

Here are three guiding questions to shape a real AI strategy:

1. What is the Real Problem?

• What is my business strategy and what are its strategic goals?
• Is there a specific challenge AI can meaningfully address?
• Are we adopting AI just because it's fashionable, or because it solves a critical pain point?
• What's obstructing the challenge? Technical? Organizational? External (e.g., market conditions or competitor hype)?

2. How Will We Approach This Challenge?

• Are we doubling down on core strengths, or chasing what Big Tech does?
• What principles will guide our AI projects (e.g., customer value first, efficiency first)?
• How do we prevent spreading resources too thin across disconnected experiments?

3. What Will We Actually Do?

• What one or two coordinated initiatives will create the most leverage beyond a symbolic pilot?
• Do we have the necessary infrastructure (data pipelines, governance, ethics review, security)?
• How will we ensure adoption and integration across the organization?

Final Thoughts

A strong AI strategy is not about following trends. It's about diagnosing the real problem, crafting a guiding policy, and executing focused, coherent actions.

If you're working on your own AI roadmap, start with these questions. Avoid fluff, stay grounded in your business goals, and make sure your AI strategy creates real, lasting value.

Let me know which of these questions resonate with you or reach out if you'd like to discuss how to build and execute AI strategy in your organization.