Strategy Function Metrics: Count Decisions, Follow the Money

Fourteen decks and a 4.3 out of 5

Picture the strategy team's annual review. The slide is full. Fourteen executive presentations. Three offsites facilitated. A refreshed five-year plan. A hundred and twenty stakeholder interviews. An internal client satisfaction score of 4.3 out of 5.

The CFO reads it twice and asks one question: “Which decisions went differently because of you?” The head of strategy starts to answer with the offsites.

Strategy function metrics are the measures a company uses to judge whether its strategy team, office of strategy management or chief strategy officer is worth what it costs. The usual ones count activity, because activity is easy to count. This page argues for a different order: count the decisions the function got made, follow the money that moved because of them, and then check, a year or two later, whether those decisions held up. Everything else is diagnostic at best.

Count the decisions. Follow the money. Then check who was right.

Why the obvious measures mislead

A strategy team produces documents, meetings and analysis. Its value is in other people's choices. That gap is why the obvious metrics mislead: they measure what the team controls, which is not what the company needs from it.

Donald Sull, Rebecca Homkes and Charles Sull describe the same trap in strategy communication: executives measure it by inputs, such as the number of emails sent or town halls hosted, rather than by the only metric that counts, how well key leaders understand what is communicated ⁴. A count of decks is the strategy team's version of a count of emails.

The deeper problem is incentive. Whatever a strategy team is measured on, it will produce more of. Measure decks and you get decks. Measure satisfaction and you get a team that avoids the uncomfortable recommendation, which is often the only one worth paying for.

Start from decision effectiveness

There is a better frame, and it comes from research on decisions rather than on strategy teams. Marcia Blenko, Michael Mankins and Paul Rogers, with colleagues at Bain & Company, surveyed executives at 760 companies in 2008, most with revenues above $1 billion, and scored each company on four dimensions: decision quality (whether decisions proved right more often than not), speed (faster or slower than competitors), yield (how well decisions were translated into action) and effort (the time, trouble and expense each key decision required) ⁵. They report that decision effectiveness and financial results correlated at a 95% confidence level or higher for every country, industry and company size in their sample, and that the most effective companies generated average total shareholder returns nearly six percentage points higher than the others ⁵.

That is a correlation from a survey of executives' own views, not proof that better decision processes cause higher returns. But it gives a strategy function the right four questions, because the function exists to improve exactly those four things for the decisions that matter most. Translated into metrics a strategy team can actually track:

• Speed: were the decisions on the executive agenda made when they were due?
• Yield: did money and people actually move once they were made?
• Quality: looking back a year or two later, did the decisions hold up against what was expected?
• Effort: how much executive time did the decisions consume, and was it time spent deciding or being briefed?

The scorecard

Six measures, grouped by the four questions. Each has a counter-measure, because any metric a team owns will eventually be managed. The definitions are our own design, not an industry standard; adjust them, but keep the pairing.

Measure 6 has a well-documented baseline. In Mankins's survey of top teams at 187 companies with market capitalizations of at least $1 billion, executives spent an average of 21 hours a month together in leadership meetings and less than three hours a month on strategic issues, and as much as 80% of their time went to issues accounting for less than 20% of long-term value ³. At ABN AMRO, splitting operating and strategy meetings took the managing board's monthly strategy time from as little as one hour to as much as ten, while total meeting time fell from 24 to 22 hours ³. That is the kind of change a strategy function can engineer, and measure.

Measures 1 and 2 have a baseline of sorts too. In a survey of 156 large companies, Mankins and Richard Steele found that those running an annual planning review by business unit reported an average of 2.5 major strategic decisions a year, against 6.1 for those reviewing strategy continuously and by issue; only 11% of the executives surveyed believed strongly that planning was worth the effort ¹. The gap between the two patterns is a survey correlation, not proof of cause, but a function that moves its company from the first pattern to the second should see it in measure 1 before it sees it anywhere else.

The lookback: the measure almost nobody runs

Measure 5 is the hardest and the most valuable. Mankins and Steele surveyed executives at 197 companies with sales above $500 million and found that, on average, strategies delivered only 63% of the financial performance they promised ². In the authors' experience, fewer than 15% of companies regularly go back and compare each unit's results with the performance forecast in its prior years' strategic plans ². Without that comparison, they argue, top managers cannot easily know whether the projections behind their investment and portfolio decisions are any good ².

A lookback fixes that, if it is run on reasoning rather than results. A good decision can turn out badly and a bad one can get lucky. So the review asks three questions about each major decision, using only what was written down at the time:

1. Did the key assumptions hold?List the two or three assumptions the decision log recorded and mark each as held, failed or unclear. This is the core of the hit rate.
2. Was the information reasonable for its time?Ask whether someone could have known better with the time and money available then. If yes, that is a process finding for the strategy team.
3. Did the review trigger fire, and did anyone act?If the decision named a signal that would reopen it, check whether the signal appeared and whether the decision was reopened.

The lookback only works if the original decision was written down with its assumptions and a review trigger, as the decision meeting checklist asks. That is the hidden benefit of the measure: to be able to score itself in two years, the function has to make decisions recordable today.

Following the money

Measure 3 is where strategy shows up in the accounts. Sull and colleagues found fewer than one-third of managers believed their organizations reallocated funds to the right places quickly enough, and only 20% said their organizations did a good job of shifting people across units to support strategic priorities ⁴. They also cite a McKinsey study in which companies that actively reallocated capital expenditure across business units achieved an average shareholder return 30% higher than the average return of companies that were slow to shift funds ⁴. That, too, is a correlation, not proof that reallocation caused the higher returns.

A simple version you can compute from your own budget: take each unit's or priority's share of total capital (or headcount) this year and last year, add up the absolute changes, and halve the total, since every point gained somewhere is lost somewhere else. If three units held 50%, 30% and 20% of capital last year and 44%, 30% and 26% this year, the changes are 6, 0 and 6 points, so 6 percentage points of capital moved. This is our arithmetic, not McKinsey's definition, and it is crude. It is still more than most strategy teams can say.

Then ask the question the number cannot: did the money move toward the stated priorities? Six points that followed the strategy is the function working. Six points that followed whichever unit lobbied hardest is the function losing.

Attribution, and the case against measuring at all

The strongest objection is attribution. The strategy team does not make the decisions; the executives do. When a decision is made on time, the CEO deserves the credit. When capital moves, finance moved it. When a lookback shows good judgment, it was the business unit's judgment. Measuring a staff function on other people's outcomes, the argument goes, is either unfair or meaningless, and the honest measure is simply whether the CEO trusts the team.

Most of that is true and none of it is a reason to stop. These measures are not a claim that the strategy team caused the results. They are a description of what the company is trying to get from having a strategy team at all: decisions made, money moved, judgment that improves over time. If those things are not happening, the team is not doing its job, whatever its share of the credit. If they are happening, the CEO can make the attribution call, and should, in writing, once a year.

CEO trust is a real measure too, but a fragile one. It rewards the team that is pleasant and quick, and it disappears with the CEO. A scorecard built on the decision log survives a change of leadership and gives the next CEO something to inherit besides a reputation.

The second objection is gaming, and it is fair. Every measure in the table can be hit cheaply: pad the agenda, shuffle capital, stop trivial initiatives. That is why each has a counter-measure, the same discipline the KPI examples page applies to any KPI. And it is why the lookback sits at the center. It is the one measure that is hard to game, because it compares today's results with what the function wrote down two years ago, when nobody knew how the story would end.

Who owns the measures

Three design choices decide whether the scorecard helps or becomes another report.

• The CEO judges; the team keeps the books. The strategy team maintains the decision agenda, the decision log and the lookbacks. The CEO or executive committee reviews the scorecard once a year and says what it means.
• Tie it to the charter. The measures belong in the team's charter, next to the decisions it serves and the work it turns down. The strategy team charter has a field for exactly this, reviewed every six months.
• Expect the numbers to look worse first. A function that starts keeping a decision log will discover decisions that were never made and assumptions nobody wrote down. Kaplan and Norton describe an office of strategy management that monitors strategic initiatives with an eye to consolidating or eliminating those not delivering strategic benefits ⁶. The first year of honest measurement usually finds more to stop than to celebrate.

For a qualitative check of where the function stands before you build the scorecard, the strategy function maturity assessment asks the same underlying question in statement form: do the executive team's decisions, and the money behind them, trace back to the function's work?

Answer the CFO's question

Monday: list the decisions your executive team made in the last twelve months that were material. For each, write whether the strategy function framed it, whether money moved afterward, and whether anyone wrote down the assumptions. That list is your first scorecard, and its gaps are your work plan.

Go back to the CFO's question. “Which decisions went differently because of you?” A strategy function that can answer it with a list, a budget line and a lookback does not need a satisfaction score. One that answers with fourteen decks has told the CFO everything.

Questions

What metrics should a corporate strategy team use?

Measures tied to decisions rather than documents: the share of decisions on the executive agenda made by their due date, time from framing to decision, the share of capital and headcount reallocated toward stated priorities, strategic initiatives stopped or re-funded on evidence, a lookback on whether past decisions' key assumptions held, and executive meeting time spent deciding rather than being briefed.

How do you measure the value of a strategy function?

Ask which decisions went differently because of it. In practice, track whether major decisions get made on time, whether money and people move after them, and whether the decisions hold up when reviewed one to two years later against what was written at the time. The strategy team keeps the records; the CEO or executive committee judges the results.

Why not measure a strategy team by internal client satisfaction?

Satisfaction tracks how pleasant the team is to work with, not whether it improves decisions. A team that tells executives what they want to hear can score well while changing nothing, and a team that delivers an unwelcome but correct recommendation can score badly. Use satisfaction as a diagnostic, never as the headline measure.

What is a decision lookback?

A decision lookback reviews major decisions 12 to 24 months after they were made, using what was recorded at the time: the options, the key assumptions and the review trigger. It asks whether the assumptions held, whether the information was reasonable for its time and whether anyone acted when the trigger fired. It judges reasoning, not just outcomes.

Where to go next

• Strategy Team Charter Template: Decisions Served, Work Refused
• Strategy Function Maturity Assessment
• What a Corporate Strategy Team Actually Does
• Decision Log
• Post-Decision Review Template
• Balanced Scorecard: Framework & How to Apply It