The strategy process: eight steps from the first question to ongoing refinement

What is a strategy process, and where does it start?

A strategy process is the ordered sequence of steps that ends in a decision about where a company wants to go and what it will give up to get there. The key thing about the word process is that it continues after the decision is made.

It doesn't start with the workshop. It starts with a question someone can no longer answer. "Will our core business still grow five years from now?" is one of those questions. Schedule the meeting before you have the question and you get a well-facilitated day with nothing at its center.

The first question in the process is organizational, not substantive: can we do this ourselves? A strategy process needs very good planning, and the people involved need to enjoy the work. Miss either one and the whole thing turns into a duty everyone sits through.

Strategy development is not planning. Planning answers how much you want to sell next year. Strategy answers with what, to whom, and why you of all people. Handle both in the same meeting and you end up with a number instead of a direction.

What are the eight steps?

Eight. They break into two halves. Steps one through four develop the strategy, steps five through eight move it into daily operations. That dividing line is the most important part of the whole overview, because that's where most processes stop.

The workflow is not a one-way street. You'll run through it several times. During analysis you notice the wrong people are at the table, so it's back to planning. During implementation it turns out a goal wasn't written in measurable terms, so it's back to focusing. That isn't a setback in the process, it is the process.

And the second half is no longer a sequence at all. In the 2022 book it still appeared as a chain of four steps. Today we describe it as an operating system: you set up launch and implement once, then execute and adjust run in a continuous loop. Underneath sits a layer that never pauses, made up of platform, performance management, and reporting routines. Plan execution as a project and you plan it with an end date that doesn't exist.

Who takes on which role?

Three roles that, in mid-sized companies, regularly land on the same people. That's exactly the problem.

• Who decides. Management, together with the division heads who have to execute it. This group stays the same throughout the whole process. Anyone who joins only for the decision meeting is deciding on material they don't know.
• Who prepares. One or two people who gather data, run interviews, and pull the documents together. That's somewhere between twenty and forty working days, spread across the process. They come on top of day-to-day business.
• Who facilitates. The person who leads the sessions, asks the uncomfortable questions, and makes sure there's a decision at the end. It shouldn't be the same person who decides.

The third role is the most common reason to bring in someone from outside. Rarely the method. In an owner-managed company, neutral facilitation from within is almost impossible to staff, because everyone in the room depends on management.

Then there's a role nobody assigns and everybody needs: someone who watches the language. When the group talks about “the market” and three people mean three different things, every analysis built on it is worthless. Clearing that up early costs half a morning and saves weeks later.

What does a realistic timeline look like?

Three to six months to an approved strategy, unevenly distributed. Take a process that starts in January: planning and analysis through mid-February, focusing in two consecutive days in early March, adapting through the end of April. Then the second half begins, and it has no end date.

What regularly blows up the plan is scheduling, not the work. Two consecutive days for six executives need six to eight weeks of lead time. If you're planning the process, set the big dates first and fill in the work around them.

Less than three months is possible, but you pay dearly for it. The analysis can be done in two weeks if someone works on it full time. What can't be accelerated is the time a leadership team needs to accept an uncomfortable finding. Those weeks appear in no project plan, and they happen anyway.

Why do so many processes end after the fourth step?

Because the first half has an end and the second doesn't. Planning, analyzing, focusing, adapting: these are projects with a conclusion, with dates, with a result you can present. Cascading and transforming are states.

On top of that, energy is at its peak at the end of the decision days, and everyone feels done. They aren't. The feeling is stronger than the insight. The bill arrives six months later, when nobody can say what became of the decisions.

Reshaping the organization also creates uncertainty. Uncertainty leads to hesitation, and sometimes to resistance. What helps against that is a community of decision-makers, multipliers, and supporters across hierarchy levels and divisions, working alongside the existing leadership line. Run the cascade through middle management alone and you lose both speed and commitment along the way.

The second reason is a performance management problem. Without a fixed rhythm, execution gets lost in status reports. The rule that prevents this is simple and still rarely followed: every report ends with a decision. Prioritize, sharpen, reallocate, or stop. A report that is merely noted has cost everyone's time and moved nothing.

What else helps is unspectacular: a fixed date in the calendar that doesn't get moved, and one person who owns it. Not a committee. One person with a name.

How often is a strategy reviewed?

Quarterly for progress, annually for validity. The difference matters. Every quarter you ask whether you're making progress. Every year you ask whether the direction is still right, and that's the more uncomfortable question, because it can call last year's work into question.

That's exactly what the eighth step is for. Adjusting means questioning and sharpening the strategy on a regular basis, and it's the only step that is never finished. A company that skips this step finds out from its results that its assumptions no longer held.

A full redevelopment is needed less often than many assume. Every three to five years, or whenever one of the assumptions the strategy rests on changes. Write those assumptions down when you approve the strategy. Then you can check later whether one of them has flipped.