Leaders who oppose a rival initiative staff its critical work with people they know are under-skilled or overloaded, so it fails on execution and nobody ever has to cancel it.
An example
Over the objections of the vice president of engineering, the vice president of purchasing at a brake and suspension supplier won approval for a supplier quality program. Engineering controlled the staffing, and it contributed four people. Their titles and recent training certificates looked right, but two had never worked on supplier audits, one was already split across three launches, and the fourth had struggled on a similar program at another plant. Engineering's best quality engineers stayed on the electric-vehicle program its vice president sponsored, which hit every gate. After two quarters the supplier program had completed nine of thirty planned audits. At the quarterly review, the steering committee looked at the two programs' milestone charts side by side and paused the supplier program for lack of results.
Why it works
Killing the initiative through a predictable capability mismatch costs the opponent nothing politically, which makes it easy; they even look cooperative. The rival initiative looks worse, because its deliverables pass through hands not equipped for them. Pursuing it becomes costlier: rework, escalations, slipping milestones, and sponsors spending capital to defend it. One variant installs weak leaders in rival units, which then need rescuing. Another sends the best people only to favored work, so the later comparison is rigged from the start. Control of assignments makes it possible, and a rival who cannot demand named talent, plus assignees who look fine on paper, keep it deniable.
The thinking behind it
George A. Akerlof Akerlof's The Market for "Lemons": Quality Uncertainty and the Market Mechanism (1970) shows how gaps in information let the better-informed side set the terms. The assigner knows who on the bench can do the work; the rival initiative's sponsor sees only titles and training certificates. Asking for named people, and the reason for each assignment, before work starts removes the information gap the move depends on.
Lee Ross In The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process (1977), Ross shows that observers underweight the situation when they explain results. When an understaffed initiative fails, the failure reads as a verdict on the initiative or its sponsor, and the staffing that produced it goes unexamined. Post-mortems that look at staffing rationale alongside results correct for this.
Warning signs
Critical roles on a rival's initiative are filled from the thin bench while stronger staff sit nearby, and the assignees look fine on paper, with titles, tenure, and recent training. A competing initiative under the same staffing authority gets the top-rated people. One executive's picks for autonomous units fail again and again, and the same executive funds each rescue. The comparison window is too short for the weaker team to recover.
How to defend yourself
Take away what the tactic depends on: the assigner's private view of who is capable.
- If you sponsor an initiative, ask for named people, with the reason for each assignment, before work starts.
- Ask for competing slates of candidates for the leaders of autonomous units, and a say for the board or an outside sponsor.
- When initiatives are compared, ask who was on each team and how they were chosen before asking how each performed.
- Escalate openly, early, and with specifics when the staffing does not match the work.
When an autonomous unit keeps failing, look at whoever selected its leaders, especially if that person gains from the failure.
How a leader shuts it down
Publish the rationale for critical assignments, and have post-mortems examine staffing alongside results. Use staffing-quality analytics across competing initiatives. Audit the selector when units keep failing under leaders one person chose. Governance bodies that ask about inputs before outputs take most of the power out of this move.