Execution and delivery
Promising What the Team Can Deliver: When the Problem Is the Job Design
Before demanding results, check whether the job has the resources, influence and support to deliver them; then sustain accountability without fear.

When someone misses a commitment, the first explanation points at the person: not enough drive, poor organization. There is another explanation that almost never gets reviewed and usually weighs more: that the job was designed to fail from the day the result was handed over.
Robert Simons, a professor at Harvard Business School, makes that case in "Designing High-Performance Jobs" (Harvard Business Review, 2005) and in his book Levers of Organization Design. His thesis: the supply of resources in every job has to match its demand, and the same goes for every department and business unit. When a good strategy is executed badly, he suggests asking whether the organization is designed to fail.
To check, he proposes four questions every manager needs answered. He calls them the spans of the job and pictures them as the sliders on an amplifier, running from narrow to wide.
First questionWhat resources they control
This is the span of control: which people, assets and information the person decides on without asking anyone.
Second questionWhat measures they are judged by
This is the span of accountability. A narrow measure, such as headcount or a line item of spending, leaves little room for trade-offs. A wide one, such as market share or customer satisfaction, leaves plenty.
Third questionWho they have to convince
This is the span of influence: how many people in other areas, who do not report to them, they need to persuade to reach their goals.
Fourth questionHow much help they can expect
This is the span of support: how much help they get when they ask for it. It depends mostly on culture and on shared goals and incentives, which is why it can hardly be adjusted for a single job.
The X test
Simons sorts the four like this: control and support are the supply of resources; accountability and influence are the demand. Draw one line joining control to support and another joining accountability to influence. If they cross to form an X, the job is balanced. If resources fall short, the strategy fails in execution; if they exceed demand, assets sit idle and economic performance suffers.
Sometimes it pays to unbalance on purpose and measure someone on more than they control. Simons calls this the entrepreneurial gap, because it forces a search for ways to deliver without commanding everything. He quotes A.G. Lafley on Meg Whitman: power shows when your circle of influence is wider than your circle of control. But it only works when the job has real influence and support. Simons describes a marketing and sales manager at a software company who answers for growth, profit and customer satisfaction although none of the areas he depends on report to him; it works because the company made customer satisfaction a shared priority.
Influence can be widened with cross-functional teams, dual reporting or stretch goals; support, more slowly, with shared objectives and incentives.
After the designAccountability without fear
With the job balanced, the second half remains. Liane Davey, in a Harvard Business Review article from February 2023, argues that you cannot hold anyone accountable: accountability is something the person feels, and the language of fear does not produce it.
Her first recommendation is to strip adjectives out of expectations. "Collaborative", "on time" or "innovative" mean different things to different people. Davey suggests replacing them with verbs and nouns: instead of "be collaborative", "ask marketing for their assessment and include it in your analysis". And answer the why, the what and the who at the start.
Then keep attention on it: make progress visible to everyone on a board and review it in one-on-ones and team meetings so problems surface early. Create enough safety for people to report difficulties, and guide with questions instead of solving the problem for them, because then the accountability becomes yours. Small, frequent corrections rather than one download of disappointment. And proportionate consequences, mostly positive: if someone keeps missing deadlines, start by adding intermediate milestones and escalate only if nothing changes.
An example
Imagine a regional manager asked to raise customer satisfaction in three countries. Technical support and logistics, which weigh most on that satisfaction, do not report to him. His accountability is wide; his control, narrow. Before demanding more, his director draws the X and sees it does not cross: he lacks influence and support.
She adjusts two things: a fortnightly committee with the heads of support and logistics, and customer satisfaction on the objectives board of all three areas. Then she sets the expectation without adjectives: "Each month you present the indicator by country, the three main causes of complaints and the action agreed with support for each one." If the indicator does not move one month, the first consequence is a weekly review.
Worth remembering
- Design first. Before blaming the person for a missed commitment, draw the X of the job.
- Supply and demand. Control and support must cover the accountability and influence being demanded.
- Expectations in verbs. Adjectives get interpreted; actions get done or not.
Closing thought
A deliverable commitment is designed before it is requested and supported after. Demanding results from a job that cannot produce them, and calling it a lack of commitment, costs more than it looks.
If you want to start this week:
1. Pick a recent missed commitment and draw the four spans of that job. 2. Rewrite one expectation that contains adjectives using verbs and nouns. 3. Decide the first, proportionate consequence if the next milestone slips.
