Strategy in the era of AI

Hello, and welcome to Future Solving, my newsletter for leaders and managers navigating what it means to lead in the era of AI.

Today, I'm going to write about strategy—what it is, how (nearly) every role is a strategy role, and the importance of strategy in the era of AI.

What is strategy?

The first time I remember hearing the word “strategy" was from a Chess Master.

I must have been five or six years old.

Learning what "strategy" meant, how to do it, and how it was different than tactics meant winning trophies.

We went to Chess tournaments nearly every Saturday.

Nearly every Saturday, I would play five games against other kids (and sometimes adults) and I would write my moves down.

Throughout the following week, my friends and I would review our games together, analyzing what our strategies were, what tactics we used, and how it went.

When I was 12, we won 1st place in the United States at the National Championship.

When I entered the corporate world, the word Strategy meant something different to me.

It still does.

So what is strategy?

The technical definitions of strategy are not helpful, as they focus on the military or define it as a plan (which it is not).

One of the prevailing thought leaders on strategy in management thinking is Rita McGrath, a Strategy Professor at Columbia Business School, who writes that to remain competitive, companies need to adopt a more dynamic view of strategy that considers rapid changes in the market and the broader ecosystem of stakeholders. This includes rethinking who their competitors are (through a lens of competitive arenas rather than industries), focusing on the jobs to be done for customers, and being agile in response to emerging threats and opportunities.

A second prevailing thought leader on strategy in management thinking is Roger Martin, former Dean of the Rotman School of Management, who says an organization can only have one strategy that requires making tough choices that are clear and opposite to what others are doing, thereby creating a unique position in the market.

I have the utmost respect for Rita and Roger's work. And if you plot Rita's, Roger's, and my views on a Venn diagram, there is a good degree of overlap, with key divergences.

My definition of strategy is:

Strategy is the process of creating and selecting decision trees of choices under conditions of uncertainty and competition, with the aim of achieving one's goals.

The key divergences are:

1. I don't believe in creating one strategy for the whole company. I do believe, however, in having a core "trunk" from which individual departments create "branches." The branches should grow out of the core strategy of the organization, but doing strategy work is how departments and teams create clarity, buy-in, and momentum.
2. "Jobs to be done" is an excellent framework for innovation and product development, but the altitude of jobs to be done is inherently tactical. Components of a strategy can be executed using the jobs to be done framework, but they are different methods for different purposes.
3. A decision tree of choices (which I call a Reason Tree in Autonomous Transformation), is purpose-built for the degree of uncertainty in a competitive market. Imagine sitting at a chess board and creating dozens of decision trees ("If I move there, then they could move x, y, and z, and if they move x, then I could move a, b, c, or d—d would set me up to control the open rank, but then they could move..." and so on. 👈 that's just one decision tree). An average chess player is carrying multiple decision trees in his or her mind at any given point, creates and considers hundreds of decision trees each game, and the depth of those decision trees is directly correlated to their degree of skill.

When you're working in a corporation, the desire is to lock in on a "strategy" and then pivot from brainstorming and strategy to execution. The issue with this approach is that the very first step of execution can already introduce enough variability that the whole decision tree must be reexamined.

(Nearly) Every role is a strategy role.

There has long been a misconception that strategy is only the job of people with "Strategy" in their title.

Here's whose role doesn't involve strategy: someone doing repetitive, transactional work.

The moment that person gets promoted into a manager position and takes accountability for achieving a specific goal with any degree of uncertainty, their job includes strategy.

If you're a manufacturing plant manager, your job includes strategy.

Social media manager? Your job includes strategy.

IT Director? Yup. Strategy.

HR director? Also strategy.

Finance director? Believe it or not, also strategy. (deep cut for my fellow Parks & Rec fans)

Most positions within organizations include strategy, but the primary focus of the education system and learning on the job is on execution with the failed premise that the strategy is stagnant or "locked" once the initiative moves from the strategy phase to the execution phase.

Strategy is one of the reasons AI projects are failing

In my book, I referenced a VentureBeat article that states that 87% of AI projects are failing. A recent Gartner paper shows that this number is not going down, with a reported 85% failure rate (and those are the failures organizations have been willing to report).

These are the key reasons I've found for why our AI projects are failing:

1. A lack of vision, which I wrote about here.
2. Tool worship (leading with technology and not with value creation), which you can read about here.
3. Data-driven decision-making (the unscientific approach to decision-making), which you can read about here.
4. Leading people as if they are cogs in a machine instead of rockstars joining a band, which you can read about here.
5. Strategy done poorly or not at all.

So where should you start?

Start with defining the vision or goal for your team or organization, then create decision trees (or Reason Trees, which you can find a framework for in Chapter 15 of Autonomous Transformation) to account for as much variability as you can imagine.

Socialize with any relevant stakeholders and document the ideas, support, and resistance they provide to your decision trees, then select the decision tree you want to move forward with based on a combination of perceived likelihood and alignment with your desired outcomes.

Then reassess as often as you can.

A new hire? Reassess.

A new competitor? Reassess.

A key team member retires? Reassess.

Ideally, reassessments will indicate no need for change in course, but it is much better to reassess often than to go several months before realizing you are headed in the wrong direction.

Due to the number of requests I've received for more resources on leveraging the Future Solving and Reason-Driven Strategy frameworks from Autonomous Transformation to build an AI Strategy that will succeed, I've created a live course that starts March 25th called AI Strategy Fundamentals. If you or your colleagues wish to join the course, I've created a $100 off promotion code ("FutureSolving") for subscribers to this newsletter. I'd be glad to see you there!

Thanks for reading,

Brian

Brian Evergreen is the author of Autonomous Transformation (Top 50 Business & Science Book of 2023 shortlisted for the Thinkers50 2023 Breakthrough Idea Award), a senior advisor on AI & strategy, Senior Fellow at The Conference Board , global keynote speaker, and teacher of AI Strategy Fundamentals on Maven .