Stakeholder Intelligence Before a Board Decision: A Working Guide

If a board is about to approve a strategic move, the quality of the decision depends heavily on what directors know about the stakeholders who will react to it. Stakeholder intelligence done well, in the weeks before the vote, is what separates a decision that holds up from one that unravels on contact with reality. This guide sets out how to run that work: what to gather, who to speak to, how to sequence it, and how to put it in front of the board.

What stakeholder intelligence actually means in this context

It is not a stakeholder map on a slide. It is primary evidence, usually from interviews, direct engagement, or structured review of recent behaviour, about how specific parties are likely to respond to the decision in question. The question is never "who are our stakeholders" but "what will the five to fifteen parties who can make or break this decision actually do when it lands."

For a board decision, that typically covers: the lead regulator and any secondary supervisors, the top clients or distribution partners whose continued support is assumed in the business case, significant shareholders, the executives and senior specialists whose retention is load-bearing, and counterparties whose consent or cooperation is required.

Sequencing the work

The single most common failure is doing this work too late. If the board pack is already drafted, the intelligence becomes a check rather than an input, and management will understandably resist findings that force a rewrite.

Aim to complete fieldwork four to six weeks before the board meeting. That gives time to test findings with the executive, revise the proposal if needed, and present a recommendation that already reflects what stakeholders think. If the decision is time pressured, compress the fieldwork but preserve the sequence: intelligence first, recommendation second.

What to actually collect

For each priority stakeholder, you want three things. First, their current position on the matter, drawn from recent direct contact, not inference. Second, the conditions under which that position would change, in either direction. Third, what they would do, not just say, if the decision went ahead as proposed.

For regulators specifically, the goal is to understand the supervisory concerns the decision will raise and to prepare to meet them properly. That means engaging early where the relationship supports it, being candid about what is proposed, and being ready to demonstrate that the firm has thought through conduct, prudential, and customer impact. The intelligence here is about readiness and credibility, not about managing perception.

Who should run it

The person running the work needs enough independence from the proposal's sponsor to report inconvenient findings. In practice that means a corporate development lead, a strategy function reporting to the CEO or chair, a board advisor, or an external firm. If the sponsor of the decision controls the intelligence gathering, directors should treat the output with caution.

What good looks like in the board pack

A short, named section, usually three to five pages, that sets out: the stakeholders assessed, how the evidence was gathered, the material findings including dissent, the implications for the proposal, and any changes already made in response. Direct quotes, anonymised where needed, carry more weight than paraphrase. Where a stakeholder's position is uncertain, say so.

What most people get wrong

The recurring errors are predictable. Relying on the relationship manager's view of a client rather than speaking to the client. Treating the regulator's last published statement as their current view. Omitting the employees whose exit would damage execution. Presenting a single consensus view when the evidence showed a split. Writing the intelligence section after the recommendation.

Your next decision

Before the next material board decision, confirm two things: who owns the stakeholder intelligence work, and whether it will complete in time to shape the recommendation. If either answer is unclear, fix that first.

Frequently Asked Questions

How many stakeholders should we actually speak to?

Usually between eight and twenty for a single board decision. Fewer than eight and you are relying on a handful of views. More than twenty and the signal gets lost. Prioritise by who can materially change the outcome.

Should the CEO see the findings before the board does?

Yes, but with a clear protocol. The executive should see findings in time to respond, but the board should see the original evidence, not a version filtered through management's reaction to it.

What if a key stakeholder refuses to engage?

Record it. Non engagement is itself intelligence, and directors should know when a view could not be obtained rather than being given an inferred one.

How do we handle regulator engagement at this stage?

Use established channels, be straightforward about what is proposed, and treat the conversation as a chance to understand supervisory expectations properly. The purpose is to meet the bar well, not to sound out how little scrutiny the firm can expect.

Can this be done internally or do we need outside help?

It can be done internally if the person running it has genuine independence from the sponsor and the access to speak candidly with external parties. Where either is doubtful, bring in an outside firm.