Fourteen Stakeholders, All "Manage Closely"
A utility's $85M smart-meter rollout has a stakeholder engagement plan listing 14 stakeholders, all classified as "manage closely," with a quarterly newsletter as the engagement mechanism for each. Three signals suggest the plan is failing. The municipal regulator, which must approve the rollout schedule, issued an information request that the program answered three weeks late. The union representing 200 field technicians, whose roles change under the program, has declined two meeting invitations and has started briefing local media on job losses. A consumer advocacy group is gaining traction in local press. A council election is five months away, and two councillors have made the rollout a campaign issue.
1. What is structurally wrong with the current stakeholder engagement plan?
Read the worked answer
Three structural faults:
- No segmentation. Classifying all 14 stakeholders as "manage closely" is the same as classifying none. Engagement effort is finite; it must be allocated by power, interest, and - critically - current attitude (supportive, neutral, resistant). The regulator and the union need fundamentally different strategies, not the same newsletter.
- Communication substituted for engagement. A quarterly newsletter is one-way. The union's grievance and the regulator's information need require dialogue, negotiation, and influence - none of which a newsletter can deliver. This is why the plan looks healthy on paper while the union declines meetings.
- No feedback loop. Nothing in the plan detects that the regulator's request went unanswered or that the union's position hardened. Engagement must be measured by stakeholder behavior and sentiment, not by the number of items distributed.
The signals described are exactly what a functioning engagement plan would have surfaced weeks earlier.
2. How should the program manager re-sequence engagement across the regulator, the union, and the political stakeholders?
Read the worked answer
Sequence by risk to the schedule and by who can actually move whom:
Regulator (first, immediately) - highest power over the rollout date, and the missed information request is a compliance failure, not a communication one. Fix the process: a named owner, a tracked request log, and a recovery meeting at senior level. Apologize once, concretely, and then demonstrate reliability.
Union (second, at sponsor level) - 200 field technicians can stop field work. A newsletter cannot address job-loss fear. Engagement must move to "involve": co-design the redeployment and reskilling plan, quantify the job impact honestly, and put it in writing. This should not be delegated to the program manager alone; the sponsor's authority is needed to make commitments.
Councillors and advocacy group (third, ongoing) - the election makes this time-bound. Brief candidates early and neutrally, publish the customer and safety benefits, and avoid anything that reads as electioneering.
Sequencing matters because regulator approval gates the schedule, and union resistance gates execution.
3. How should the program manager measure whether engagement is actually improving?
Read the worked answer
Measure behavior and position, not activity:
- Stance tracking - record each key stakeholder's position (supportive/neutral/resistant) at each engagement, and whether it is moving. The union going from "declined meetings" to "attending with conditions" is progress; another newsletter sent is not.
- Response metrics - regulator requests answered within agreed turnaround; escalations raised by stakeholders rather than discovered by the program.
- Commitment evidence - signed agreements, attendance by decision-makers, joint documents produced with the union.
- Escalation count and severity - rising escalations to the sponsor or governing body are a leading indicator of engagement failure.
- Media and political sentiment - tracking coverage is cruder but relevant when an election is imminent.
Report these to the governing body alongside schedule and cost. A stakeholder plan with no measurement is a distribution list.