Testing Internal Assumptions About Stakeholders Before You Commit

Start with what you actually believe

Before you can test an assumption, you have to name it. Most executive teams skip this. They talk about stakeholder support in general terms ("the regulator is comfortable," "our top clients want this," "the board is aligned") without ever writing down the specific belief that underpins the decision.

Pull the team into a room and force the question: what exactly do we believe each material stakeholder thinks, and what would they do if we proceeded? Write it in one sentence per stakeholder. If you cannot, you do not have a position, you have a feeling.

Good looks like: "We believe the PRA will accept a phased capital treatment because informal signals from the supervisor in Q2 suggested flexibility on transition." Bad looks like: "The regulator is supportive."

Trace each belief back to its source

For every assumption, ask two questions: who told us this, and when? You are looking for three failure patterns.

First, stale evidence. A conversation from eighteen months ago with a stakeholder who has since changed roles, or whose organisation has shifted priorities, is not a current data point.

Second, single-source risk. If your view of a stakeholder rests on one relationship holder inside your firm, you are exposed. That person may be filtering, inflating access, or hearing what they want to hear.

Third, inference dressed as fact. "They didn't push back in the last meeting" is not the same as "they support it." Silence gets read as consent far more often than it should be.

Mark each assumption with a confidence grade based on evidence quality, not on how strongly people feel about it internally.

Design tests proportionate to the stake

You do not need to validate every assumption to the same depth. Rank them by two factors: how central the assumption is to the decision working, and how costly it would be to discover you were wrong after committing.

For the assumptions in the top-right quadrant, use direct tests:

• Structured bilateral conversations with the stakeholder, framed as consultation not announcement. Ask open questions about their concerns and criteria, not "are you supportive?"
• Third-party soundings through advisers, non-executives, or industry contacts who can hear things your firm cannot.
• Written responses to a specific proposition. What people say in writing is more considered than what they say in a corridor.
• Behavioural signals: how has this stakeholder actually responded to comparable moves by peer firms?

For lower-stake assumptions, desk research and internal challenge sessions may be enough.

Run a proper red team

Appoint two or three people, ideally including someone outside the deal team, to argue the opposite case. Their job is to build the strongest possible version of "the stakeholders do not support this and here is why." Give them access to the evidence base and a week.

What you are looking for is not whether they win the argument. You are looking for the moment where the sponsoring team cannot answer a specific challenge without saying "we'd need to check." Every one of those is an untested assumption.

Watch for the political filter

By the time stakeholder intelligence reaches the executive committee, it has usually been softened. Relationship owners have incentives to report warmth, deal sponsors have incentives to report momentum, and no one wants to be the person who slowed the decision.

Two counters work. First, ask for verbatim quotes, not summaries. Second, ask what would have to be true for this stakeholder to actively oppose us, and whether any of those conditions are already present.

Know when you have enough

You will never eliminate assumption risk. The question is whether you have enough validated evidence on the load-bearing beliefs to commit. A practical threshold: for every material stakeholder, you can point to at least two independent, recent sources supporting your position, and you have a specific answer to what would change their view.

If you cannot meet that bar on a stakeholder whose opposition would derail the decision, do not proceed. Delay, test further, or reshape the proposition.

Next step

Before your next major committee paper, take thirty minutes to list every stakeholder assumption embedded in the recommendation, grade the evidence behind each one, and identify the two weakest. Fix those before the paper goes in, not after.