Validating Stakeholder Assumptions Before a Regulatory Approval Decision

Start with the assumptions, not the stakeholders

Most validation exercises fail because they begin with a stakeholder map. That is the wrong starting point. Begin with a written list of the load-bearing assumptions inside your approval case: what the regulator wants, what the board will tolerate, what the business can deliver, what competitors will do, what consumer groups will say publicly. If an assumption being wrong would change your submission, your timing, or your remediation commitments, it goes on the list.

Keep it tight. Ten to fifteen assumptions is usually right. More than that and you are describing the world rather than testing your case.

For each assumption, write down two things: the evidence you currently have, and what would have to be true for it to be wrong. The second column is where blind spots live.

Sort assumptions by consequence, not confidence

The instinct is to test the assumptions you are least sure about. Resist it. Test the assumptions where being wrong would cause the most damage, regardless of how confident you feel. Confidence is often a function of who told you, not what is true.

A useful sort: which assumptions, if invalidated 48 hours before submission, would force you to pause? Those are your priority one. Everything else can wait.

Run parallel validation, not sequential

The fastest validation is not a single well-designed process. It is three or four cheaper, less precise ones running at the same time, triangulated at the end.

In practice:

• Adversarial internal review. Pick two or three people inside the firm who were not part of building the case. Give them the assumption list and ask them to argue the opposite. Not critique. Argue. One hour each.
• Quiet external soundings. Use trusted intermediaries, former regulators, sector counsel, specialist advisers, to test specific assumptions about regulator posture. Frame as hypotheticals. Avoid your own name where you can.
• Field signal check. Talk to the people who deal with the affected customers, intermediaries, or counterparties weekly. Not their managers. Ask what they are hearing that contradicts the official line.
• Document the silence. What are you not hearing from stakeholders you would expect to hear from? Silence from a usually vocal consumer group, a quiet regulator desk officer, a board member who has stopped asking questions, these are signals.

Running these in parallel takes five to seven working days. Sequentially, it takes a month.

The blind spot test most people skip

Ask this question explicitly: who benefits if our approval is delayed or refused, and what are they doing about it? Competitors, activist shareholders, politicians with adjacent agendas, internal factions who lost the argument. You do not need to be paranoid. You do need to have considered it on paper.

The related question: which stakeholder have we stopped talking to because the conversation became uncomfortable? That stakeholder is almost certainly the one whose position has shifted most.

What good triangulation looks like

After the parallel streams report back, you are looking for three patterns:

1. Convergence. Multiple independent sources confirm an assumption. Treat as validated, but note who would dissent.
2. Contradiction. Sources disagree. This is the most valuable output. Do not average them. Work out why they disagree, because the reason usually points to the real issue.
3. Absence. Nobody has a view, or the views are vague. This is usually a sign that the assumption is untested in the market, not that it is uncontroversial.

The common mistake is to weight the most senior or most familiar source highest. Weight the source closest to the actual decision or behaviour you are trying to predict.

What to do with what you find

You will usually find two or three assumptions that need reworking, one stakeholder position that has moved without you noticing, and one risk that nobody had named. That is a successful exercise. If you find nothing, your validation was not adversarial enough, run it again with different people.

Document the changes to your submission, the assumptions you consciously chose not to act on, and why. That record is what makes the decision defensible later.

Your next move

Before the end of this week, write the assumption list. Not the stakeholder map, the assumption list. If you cannot get it onto one page, you do not yet know what you are betting the approval on.