An engineering manager at a large petrochemical site told me, without much emotion, that he had started three leadership programmes in six years and finished none. He is not disengaged — he mentors two graduates and runs a genuinely good section. The programmes just kept losing to Tuesday.
Completion rates for senior technical cohorts are consistently worse than for corporate functions, and the reasons are structural rather than motivational.
The scheduling assumption is wrong
Most executive programmes are designed around a professional who controls their diary two weeks out. A plant manager does not. A trip, an outage, a Section 54 stoppage, a contractor incident — any of these vaporise a week's plan with no notice.
Programmes built on fixed monthly contact sessions punish exactly the people they were designed for. Miss two and you are behind; miss three and you disengage entirely because catching up looks worse than quitting.
The designs that survive contact with an operating site have redundancy built in. Recorded core content, two possible dates for each contact session, cohort work that can absorb one missing member, and an explicit re-entry path rather than a quiet drop-off.
Generic content loses credibility in the first hour
Senior technical people are unusually intolerant of vagueness, and rightly so. A case study about a retail turnaround in Ohio, presented to eighteen mining engineers, does measurable damage in the first session. Once the room decides the material is not serious, attendance becomes discretionary.
What holds a technical cohort is content that respects their existing expertise and extends it into unfamiliar territory — capital allocation, contract law, regulatory negotiation, dealing with a hostile community meeting, managing a union relationship through a restructure. Hard, specific, and adjacent to what they already do well.
I would rather run six sessions on real problems the cohort brings than twelve on a competency framework.
The unspoken transition
The genuine difficulty in this population is not learning management theory. It is the identity shift from being the person who solves the technical problem to being the person who is accountable for others solving it.
Almost every newly promoted engineering manager over-functions technically for the first eighteen months. They review drawings they should delegate, attend fault-finding they should leave alone, and end up as a bottleneck while their actual job — resourcing, prioritisation, developing the section, managing upwards — gets done badly in the gaps.
Naming that pattern explicitly in a room of peers does more than any module. The relief when someone says it out loud is visible.
Sponsorship is the strongest predictor
Across the programmes I have run, the single best predictor of completion is not seniority, prior education or stated motivation. It is whether the delegate's own manager was involved.
Involved means a conversation before enrolment about why this person and what should change, a check-in midway, and a role in whatever the delegate produces. Where that exists, completion rates run above 90%. Where enrolment came from an HR nomination list the delegate first heard about by email, expect a third to fall away.
Design principles I now insist on
- Duration of six to nine months. Shorter does not change behaviour; longer than a year outlives the organisational context it was designed for.
- A live business project with a real sponsor and a real decision at the end. Not a case study, not a presentation to the cohort.
- Cohort size of twelve to sixteen — enough diversity of experience, small enough that absence is noticed.
- Mixed seniority across functions. The most valuable conversations in my last cohort were between a maintenance manager and a finance manager discovering how the other one thought about deferred capex.
- An explicit contract with the line manager at the start, in writing, covering time release.
Measure the decision, not the reaction
Happy sheets at the end of a session tell you about the catering. If you want to know whether a programme worked, look eighteen months out at whether the participants are making different decisions — delegating more, escalating better, allocating capital differently, retaining their people at a higher rate.
That is harder to collect and considerably more honest. It also, in my experience, produces a much shorter list of programmes worth continuing to fund.