Strategies often stall between functions, where one team depends on another to deliver its part. Research on strategy execution finds that managers can usually rely on their own reporting line, and far less often on colleagues elsewhere. Before a launch, check the handoffs the plan relies on, who owns each one, and which priorities have the resources to carry them.
Where does a strategy usually stall?
More often between units than inside them. A strategy is approved as one decision, and then carried by many teams whose work depends on each other. Those dependencies are where commitments slip, because a team can be held to account for its own targets while the handoff it needs from another team has no clear owner.
The best-known research on this points the same way. In Sull, Homkes and Sull's study of nearly 8,000 managers in more than 250 companies, managers could rely on their own line far more than on other functions, and the authors conclude that alignment up and down the hierarchy is generally sound. The weak joint is coordination across it.
This changes what a leadership team should look at. The org chart shows the lines that already work. The places a new strategy will strain are the interfaces it adds or loads more heavily: a product team that now depends on regional sales, a shared service that now carries a growth target, a decision that now needs two functions to agree.
Why is the first explanation for a stall so often wrong?
Because friction arrives looking like a people problem. A slow decision reads as a leader without pace, a missed milestone as weak discipline, conflict as a team out of alignment. Each is a hypothesis. The same signs are produced by split authority, late evidence, or a plan that rewards incompatible outcomes across functions.
Treating the first explanation as the finding leads to the most expensive move first: replacing a leader, or launching a culture programme, when the cause sits in how the work is arranged. The discipline is to hold several explanations open and look for the evidence that separates them before acting on any.
What should a leadership team check before launching a strategy?
Five things the plan depends on: that people know which strategy now applies, what to choose and what to drop, what their own contribution is, who decides and who delivers, and which interfaces carry the result. The plan tends to give way at the weakest of the five in Atlas's reading, and dependency is the one research most often finds weak.
In practice that means four questions per priority. Which teams must hand work to each other for this priority to land? Who owns each handoff end to end? Does the priority have the money and the people it needs, or is it one of several competing for the same capacity? And what would someone need to decide differently on Monday for the strategy to be real?
That final question is a useful check of translation. A strategy that has been announced but has not changed a single decision is still a document.
Does it matter how many priorities a strategy carries?
Yes, because priorities compete for the same people. Sull and colleagues found that only 11% of managers believe all their company's strategic priorities have the financial and human resources they need. When everything is a priority, teams choose for themselves, and they choose differently.
The fix is rarely more resource. It is an explicit list of what the strategy stops doing, agreed at the top and visible to the teams that would otherwise keep doing it. Dropping work is a decision, and it needs an owner like any other.
What are the early signals that a strategy is stalling?
Repeated escalation of the same kind of decision, work arriving late at the same handoff, and a growing amount of informal coordination that keeps things moving despite the structure. Each shows people absorbing a gap between what the plan asks and what the organisation currently supports.
These signals usually appear before a missed number. By the time a milestone slips in a board pack, the teams involved have usually been compensating for a while, through workarounds, extra meetings and personal effort. That effort is real cost, and it is finite.
Where this stops
This note describes where execution tends to strain inside an organisation. It says nothing about whether the strategy itself is sound, whether the market will behave as planned, or whether the numbers behind it hold. A clean read of the handoffs is compatible with a strategy that fails for commercial reasons, and the research cited here is based on managers' own answers, which record what people report rather than what they do.
How Atlas reads it
Atlas reads the organisation behind a new strategy in five layers: the formal design, who really owns what, the dependencies the strategy relies on, how teams decide and escalate under pressure, and what people report compared with what the work shows. No single layer is the organisation. The reading is set against what the strategy now demands.
The result is an Execution Forecast: where the plan is likely to stall, what to change first, and who owns it, with the evidence for each judgement and what argues against it. The ratings describe the arrangement around a person and leave the person out of the judgement. The first recommended move is usually structural: clarify a mandate, resequence a priority, or assign an owner to an interface.
Sources
- Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
- Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
- Gallup, The Great Detachment: Why Employees Feel Stuck (2024)
Questions
Research points to coordination across functions and units rather than alignment within them. Sull, Homkes and Sull found that only 9% of managers can rely on colleagues in other units all the time. Plans depend on those handoffs, and a handoff with no clear owner is where commitments often slip first.
Sometimes, but it is rarely the first thing to conclude. Slow decisions and missed milestones are also produced by split authority, late evidence and competing priorities. The useful move is to check those structural explanations before a judgement about a leader, because the structural fixes are cheaper and faster to try.
Few enough that each has the money and people it needs. Only 11% of managers in Sull's research believed all their priorities were properly resourced. A shorter list with an explicit set of things the strategy stops doing usually executes better than a long list that teams have to ration themselves.
Before the strategy is launched, and again when something material changes: new ownership, a new operating model, new pressure, a new dependency or new evidence. Checking before launch keeps the cheaper options open, such as reassigning ownership or resequencing work, before the organisation has started compensating for the gap.
It means that people can say what they will decide differently because of it. A strategy is translated when its priorities have changed specific decisions, owners and handoffs in the teams that carry it. Until then it remains an announcement, however clearly it has been communicated.