The failures are consistently organisational rather than strategic, which is why better strategy documents have not reduced the rate.
Why it matters when the plan changes
Failure rates for large change programmes have been quoted for thirty years and the causes reported have barely changed, which tells you the causes are structural rather than a matter of competence. Sull, Homkes and Sull found that only 11% of managers believe their company's strategic priorities carry the resources needed for success, and only 9% say they can rely on colleagues in other functions all the time, well below the reliance managers report within their own line. Organisations keep improving the parts they can control easily, the plan and the communication, and keep leaving the dependency and ownership questions that actually decide whether execution lands.
The tension is that the well-known causes are not the ones that get budget. Clarifying who decides and sequencing dependencies are unglamorous, cheap and owned by nobody in particular. Programme governance, dashboards and communication are visible, expensive and easy to point at in a steering meeting. The money follows visibility rather than the failure literature.
In practice
A programme is designed with a strong sponsor, a detailed roadmap and a communication plan. It fails on three dependencies between functions that nobody owned end to end. Every post mortem in the sector describes this pattern. The next programme in the same company is designed the same way, with a better dashboard.
Evidence
Most managers do not believe their company's strategic priorities are properly resourced, which is among the most consistent findings in this literature.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)Reliance on colleagues in other functions is far weaker than within a reporting line, which is where large programmes break.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
This body of work identifies recurring causes across many programmes; it does not predict whether a specific programme will fail, or which of the causes will be decisive in a given case. It is diagnostic after the fact more often than it is predictive before the programme starts, and it says little about how fast a given organisation can close a dependency gap once found.
Questions
Priorities that are stated but not resourced, decisions the plan depends on that nobody owns, dependencies between units that nobody sequences, measures and incentives still pointing at the previous model, and leadership attention moving to the next thing before this one has landed.
Rates vary by definition and source and should be treated with care. What has stayed stable is the pattern of causes: Harvard Business Review's 2015 study by Sull, Homkes and Sull found only 11% of managers believe priorities are properly resourced, a finding echoed across decades and sectors. The rate is quotable; the cause is useful.
Because they sit in the places that are hardest to see and least rewarding to fix. Ownership and sequencing live between functions, where nobody is measured, while plans and communications live with people who are. The effort goes where the accountability is.
Marginally. The plan is rarely the failure point; what fails is the organisation meeting it. More detailed planning improves the artefact that was not the problem, which is why decades of improved planning practice have not moved the reported failure rate.
Naming the decisions a plan depends on and giving each one an owner, sequencing the interfaces it runs through, and checking that stated priorities carry resources; Sull, Homkes and Sull found only 9% of managers can rely on other functions all the time, which is why cross-unit dependencies need explicit sequencing rather than assumed goodwill.