Ask a leadership team for their priorities and you rarely get three. You get fourteen, plus two transformation programmes and an innovation initiative. Everything is important, which is another way of saying that nothing is.
The research is blunt about where the time goes. McKinsey’s global survey finds managers spend 37 percent of their time on decision making, rate well over half of it as ineffectively used, and put much of the blame on overloaded agendas: when everything is declared a priority, every decision competes with every other one for attention[U]. Fewer priorities is not a slogan. It is the precondition for decision quality.
Priorities are decisions in disguise
A priority list is not a communication artifact. It is a stack of decisions: what gets resources, what gets attention, and, most importantly, what does not. A list with fourteen priorities is fourteen deferred decisions. The organisation reads it correctly: keep doing everything, and let the calendar decide what actually matters.
The tell is simple. If nothing was stopped when the priorities were set, no decision was made. Something was merely written down.
Why lists grow
Priority lists grow for human reasons, not analytical ones. Cutting a priority means telling someone their project does not matter most. Keeping the list long avoids that conversation. It also spreads accountability so thin that no single failure can be traced to a single choice. Long lists are comfortable. That is exactly what is wrong with them.
The trade-off test
A real priority passes a simple test: you can name what you gave up for it. Not in the abstract, but concretely. We do this, therefore we do not do that, and here is who agreed to the loss. If nobody can name the loss, the priority has not been decided. It has been announced.
Working through this test with a leadership team is uncomfortable for about ninety minutes. Then it is liberating. Most teams discover that of their fourteen priorities, three carry real consequence, four are maintenance dressed up as strategy, and the rest exist because removing them was never put to a decision.
| Announced priority | Decided priority |
|---|---|
| Added to the list | Something else was removed |
| Nobody owns the loss | The loss has a name and a signature |
| Survives every review unchanged | Reviewed against results |
| Fourteen coexist peacefully | Three compete for real resources |
The practical takeaway
Take your current priority list and write next to each item what was explicitly given up for it. Where the column stays empty, you have found the work: not more analysis, but one honest trade-off conversation per line.
The AI portfolio has the same disease
The clearest current example of too many flags is the average AI initiative list. Gartner attributes its prediction: more than 40 percent of agentic AI projects canceled by end of 2027, largely to early experiments driven by hype and misapplied before anyone asked what business value they serve[GA]. That is a prioritisation failure, not a technology failure. A portfolio of two or three AI cases with owners and success criteria will outperform a portfolio of twenty pilots every single time, the same arithmetic as everywhere else in strategy.
What I see in the field
Ask a leadership team for their priorities and count the answers: past five, the word has lost its meaning. The strongest teams I work with do something uncomfortable once a quarter. They name what they are explicitly not doing, in writing. That list creates more speed than any new initiative.
Sharper trade-offs are a discipline you can install: a priority map with a decide-now list, a park list, and a kill list, each with an owner. We build it with you and keep it honest quarter after quarter.
Personalised answer within 48 hours.
Sources
- McKinsey & Company, Decision Making in the Age of Urgency (global survey), 2019. mckinsey.com
- Gartner, Press Release: Over 40 Percent of Agentic AI Projects Will Be Canceled by End of 2027, June 2025. gartner.com
