What the heck is Strategy?

Strategy is one of the most overused and overloaded words in management, let's ponder a bit on how to define it and, more importantly, differentiate good strategy from bad one.

During my sabbatical, I found myself wondering on the concept of “Strategy” quite a bit.

At Bolt we talk about Product Strategy, Vertical Strategy, Market Strategy, Organisational Strategy and many other strategies. But when trying to define what Strategy is and how it differentiates from other things, I always tend to get vague and bring comparative examples rather than specific explanations.

Even more importantly, a common piece of feedback I give to my growing managers is to develop their strategic thinking. But again when asked what exactly is strategic thinking and how to develop it, I have trouble giving specific guidance. It’s almost like an obscenity — I know it, when I see it.

Doing research for the blog post, I looked up the original definition of the word in the military context:

The art of planning and directing overall military operations and movements in a war or battle.

It’s not very helpful, as it points to scale rather than complexity, but in business there is often a hierarchy of strategies and one’s tactics can be other’s strategy.

Wikipedia definition is a little better:

Strategy is a general plan to achieve one or more long-term or overall goals under conditions of uncertainty.

The reference to long-term and uncertainty is very important, but it’s still too vague to be really helpful.

So what’s strategy in business context, and more importantly what differentiates good strategy from bad strategy?

The best analogy I found that is helpful to understand is playing chess. To be successful in chess you need to be able to predict the evolution of the game over the long term. But humans are not computers, so we can’t just calculate all possible moves — we need to build a predictive model that yields a best possible move knowing the state of the board and the opponent playing style.

To build that model, chess players accumulate a great deal of knowledge. They group moves into openings, defences, offences, endgames. They select the most likely opponent moves to focus on. They split the board into local engagements, while not forgetting about the overall game. For me, this is exactly what I do in business, so let me explain the analogy and the concept step by step.

Before we try to define strategy, let’s try to define strategic thinking analogous to the chess grandmaster in the example above:

• Depth. You need to fully understand the inner-workings of the domain under your control. This can include KPIs, unit economics, value modelling, organisations, processes and so on — every single thing that falls under your control. This is the easiest skill to develop and relies on tenacity.

• Breadth. You need to have a solid understanding of the relevant external forces, that influence your domain, but that you don't control directly. This can include market forces, user behaviour, other departments or functions, larger business goals that you need to support and so on. This skill relies on curiosity and is harder to build.

• Abstraction. You need to pick the relevant level of internal and external forces to deal with in each situation — like a chess player focusing on the likely group of moves you need to build useful abstractions over the forces at play and pick them in each situation. This skill is harder yet and relies on intelligence.

• Creativity. The best moves in chess are not based on established rules, but on breaking them in ways unexpected for the opponent. Similarly, the best strategic thinking is not linear and takes leaps of creativity that others won't come up with.

Now let’s talk about the differences from the chess game. One of the most pervasive issues in the real world, where you don’t control the board, is uncertainty. Predicting the future over a period of 5-10 years is largely an exercise in futility. Yet I commonly see 5 year plans with 2 numbers after comma filled in.

It’s easier to invent the future than to predict it. — Alan Kay

So when we talk about strategy, or any long term planning, we should capture that uncertainty. Which is why I dislike the Wikipedia definition, as the word plan invokes a fixed world, not an ever-shifting one. The word that I often use instead is bet.

Bet is something that you think is right to the best of your knowledge, but may just turn out wrong. You can have high and low confidence bets. Your biggest, highest confidence, long term bets become a part of your strategy.

Let's now try to define strategy:

A complete set of high confidence high impact bets on the long term evolution of the domain based on the deep understanding of the domain and a sufficiently broad understanding of the external forces affecting the domain that can be readily communicated and converted into execution plans.

Here domain is anything currently under your control, a metasyntactic variable if you will.

This definition is a little awkward on the other extreme — some things don’t have uncertainty and just need to be done, like transitioning to a new tax code. But importantly it incorporates the way to tell good strategy from bad one — you test the assumptions that lead to the final construct:

• Depth. One of the best tests to ensure a sufficient understanding of the domain is the classical 5 Whys. Basically, keep asking for assumptions behind assumptions to make sure that people have taken everything relevant into account.

• Breadth. The best way to test breadth, is to invite experts from other domains (that act as external forces) to review the assumptions and cross-check that there are no unforeseen circumstances. A good indicator is also expert participation in the development of the strategy.

• Completeness. This one is the hardest to fully check, but a good starting point is reviewing the bets that didn’t make it into the strategy and testing them same way. Most commonly the issue isn’t the unknown unknown, but rather underestimation of impact or relevance.

• Readiness. Strategy isn’t a set of bullet points, but a long-term execution direction for the domain — the bets should be well communicated and execution plans prepared. Very often issues with ability to explain the bet to all parties and prepare plans point to issues with the bet itself.

Another note is that you may have bets with enormous impact, but not enough confidence to include in your strategy. In which case, I kid you not, you can include it as a part of meta-bet that raises the confidence of other bets, e.g. “Research product expansion opportunities to support market size growth” is a reasonable enough meta-bet.

Most importantly, understand that your strategy is not a final plan set a stone, but just a series of your best bets, which need to be continuously tested, evaluated, added, removed and refined to build the best possible future for your business.