Outgrace
OutgraceThe Leadership Magazine, home
StrategyHow-to guide

How to write a three-year strategy a board can hold you to

A strategy becomes testable when it has a few milestones, one named owner for each, and a fixed date for the board to check the evidence.

A detailed view of a chess board setup in a professional meeting room, highlighting strategy and planning.
Photo: Alex Urezkov / Pexels

Most three-year strategies fail the board in the same place. The document describes a direction well enough, but nobody can say in March whether the company is on it, who is responsible for the next step, or what would make the directors change course. A strategy a board can hold you to has a short list of milestones, a named owner for each, and a date when someone outside the executive team looks at the evidence. It is not a longer plan. It is a plan that can be shown to be wrong, which is the only kind a board can test.

A note on sources before the method. The sources we could check for this guide cover shareholder proposals, the structure of voting rights and the accountability of boards in general terms. None of them sets out how to write a strategy, and none names a national corporate governance code or reports board-practice research on strategy review. What follows is editorial judgement, drawn from how boards tend to succeed and fail with multi-year plans. Where a point rests on a source, the source is named. Where it does not, treat it as a working method to test against your own articles of association and any code your listing rules refer to.

Before you start

Three things have to be true before the first milestone is drafted. Without them the rest of the work produces a more detailed version of the same vague document. First, the board and the executive team must agree what success looks like in year three in no more than three numbers. Second, the company needs a baseline for each number, measured the same way every quarter. Third, the board needs a committee or a standing agenda item that will own the review, so the strategy does not return to the board only when it is in trouble.

  1. Step 1

    Write the year-three test

    Describe the company as it should look at the end of year three, in three to five sentences a director could repeat from memory. Then turn each sentence into something that could be counted or observed. If a statement cannot be checked against a figure, a contract or a dated decision, it belongs in the vision statement, not the strategy. Usually what goes wrong here is a list of aspirations that all sound equally important.

  2. Step 2

    Turn each test into a milestone

    For each year-three condition, set a milestone for the end of each year and a threshold for each quarter. A threshold is a number the board can compare against the actual figure. Milestones tend to fail when they measure activity, such as launches or hires, rather than results, such as revenue retained from a product or the share of plants running to the new schedule. Keep activity measures only where they are the leading signal for a result.

  3. Step 3

    Name one owner per milestone

    Each milestone needs one executive who answers for it, by name and title. A committee or a function can contribute, but it cannot be the owner. Shared ownership usually means that when the number is missed, each executive can point to a colleague. Where the owner does not control the inputs, say so in writing and name the person who does. The board should be able to ask a single person why a number moved.

  4. Step 4

    Set the quarterly review

    Put the review on the board calendar for every quarter of the three years, with a fixed agenda and a fixed evidence pack. The pack should show each milestone, its threshold, the actual figure, the owner's explanation and the next action with a date. Directors can test a review only if they see the same format each time. The usual failure is a slide deck that changes shape every quarter, which makes the trend impossible to read.

  5. Step 5

    Draw the line on decision rights

    List the decisions management may take alone during the plan, the decisions that need board approval, and the decisions that need a board committee first. Put dollar or euro limits and time limits against each. Martin Lipton of Wachtell Lipton argues, in a piece on the Harvard Law School Forum, that one-share-one-vote lets shareholders hold boards and management accountable. Inside a company, the same logic applies: a board can only hold management to a milestone when both sides agree who decide

  6. Step 6

    Write the triggers for change

    Agree in advance the conditions under which the plan will be revised: a milestone missed by a set margin for two consecutive quarters, a material change in a key input, or a decision by the board itself. Record them in the strategy document. When a trigger is met, the review becomes a decision about the plan rather than a debate about whether anyone failed. Without written triggers, boards tend to revise plans late, after the spending has been committed.

  7. Step 7

    Run one quarter before the board sees it

    Before the first formal review, ask the owners to produce the evidence pack for one quarter using the real data. The exercise will show which measures cannot be collected on time, which thresholds are unclear, and which owners do not yet know they are owners. Fix these before the strategy is presented. A dry run costs a few weeks and saves the first board review from becoming a discussion of method.

What usually goes wrong

The most common mistake is a set of milestones that measure effort. A company that commits to opening new markets, hiring a regional team and launching a product can report progress on every one of those items and still have no evidence that the strategy is working. The board should ask what the milestone would look like if the strategy failed, and then check whether the measure would show it. A milestone that cannot fail is not yet a milestone.

The second mistake is shared ownership. In family-owned and founder-led businesses in particular, the owner of a milestone is often the person who can make the decision informally, which is not the same as the person who is accountable on paper. Cravath, in its analysis on the Harvard Law School Forum of multiclass structures, notes that investors have become more receptive to founder control when it is paired with an empowered board and a time-based or event-based sunset provision. The same discipline applies to a strategy. Where one person holds outsize influence, the board needs written decision rights and a date by which those rights are reviewed.

The third mistake is a quarterly review that becomes a status update. Directors receive a long narrative, nod at the progress and move on. The fix is to make each review end with a decision: continue, change a milestone, or invoke a trigger. If the meeting cannot end with one of the three, the evidence pack was not designed to support a decision. Ask the chair to read the minutes of the last three reviews and check how many of them recorded a decision of any kind.

The fourth mistake is treating the plan as a promise rather than a hypothesis. Madison Condon, in a 2020 article in the Washington Law Review, argues that diversified investors may internalise some effects that a single firm would ignore. A board that tests its strategy against a wider set of effects, such as supply, regulation or customer concentration, is more likely to see a problem before it reaches the milestone. The strategy should say what the company believes and what evidence would show it is wrong.

How many milestones should a three-year strategy have?

Three to five is a workable number. Each milestone needs an owner and a threshold reviewed every quarter, and a longer list usually means that some of them have no real owner. If a candidate milestone cannot be placed under one of the year-three conditions, it probably belongs in an operating plan rather than the strategy.

What if the market changes faster than three years?

Keep the three-year horizon for the milestones and review the assumptions every year. The assumptions are the inputs that would change the plan, such as prices, regulation or a key customer's behaviour. Write them down in the strategy, so that the board can see when a change in the world, rather than a failure in execution, is the reason a milestone has moved.

Who should own a milestone when two executives both have a stake in it?

Choose the one who controls the decision that most affects the number, and name the other as a contributor in writing. If the board cannot tell from the paperwork who answers for a result, the milestone is not ready. Shared credit is fine; shared accountability is where results go missing.

Does the law require a board to set a three-year strategy?

The sources we checked for this guide do not establish such a requirement, and we have not stated one. Whether your company must set out a multi-year plan depends on its articles, its listing rules and the code it reports against. Put that question to your general counsel or company secretary before the first board review.

What should happen when a trigger is met?

The board should hold a separate meeting or agenda item to decide the response, and the decision should be minuted with its reasons. The response might be to change a milestone, reallocate spending or stop a workstream. What matters is that the decision is taken at the time the trigger is met, not after another quarter of spending has been committed.

This week, take the current strategy document and count the milestones it contains. For each one, write the owner's name, the quarter-end threshold and the next date on which the board will see the number. Where any of the three is missing, that milestone is not yet testable. Send the list to the chair before the next board pack is assembled, and ask for two hours of the executive team's time to close the gaps. The strategy will be tested by the board whether or not these items are filled in; it is better to be the one who wrote the test.

Newsletter

The Briefing

One email a week: the leaders we profiled, the ideas worth stealing, and nothing else.

Weekly, free, one click to leave. Read our privacy policy.