Introduction to Stakeholder Engagement.

For the purposes of research and innovation, a stakeholder is anyone whose life is affected by the work you do, or who has a defensible interest in how it is done. That definition is broader than the common reading, and it is broader on purpose.

It is worth being specific. Three useful cuts:

• Direct or indirect. Direct stakeholders interact with your work or its outputs themselves. Indirect stakeholders feel the consequences without ever encountering the work.
• Affected or interested. Affected stakeholders carry a real cost or benefit from the outcome. Interested stakeholders have a professional, regulatory or representational reason to follow the work.
• Voiced or under-voiced. Some stakeholders already have a seat at every table. Others would not be invited unless someone deliberately invited them.

The single most common mistake in stakeholder thinking is to list only the people who are already at the table. The whole point of the exercise is to surface the ones who are not.

Once you have a list of stakeholders, the next move is to sort them. The two-axis map is the simplest tool that still produces decent decisions.

The horizontal axis is interest: how much does this stakeholder care about what you decide? The vertical axis is power: how much can they actually shape what happens, either by funding, by approving, by adopting, or by blocking?

Plot each stakeholder on the grid. Four quadrants fall out:

• High interest, high power. Manage closely. These are the people the project succeeds or fails with.
• High interest, low power. Keep informed. They care, but cannot move things on their own. They are often your best evidence base.
• Low interest, high power. Keep satisfied. Quiet now, but able to derail a project later if they feel surprised.
• Low interest, low power. Monitor. Light touch.

This map is not the answer. It is a way of making the implicit explicit so a team can have a useful conversation about who they are actually planning to engage and who they are tacitly planning to ignore.

Most stakeholder conversations fail because the team running them did not decide, in advance, what would count as a successful one. Five questions to settle before you walk in the room:

1. Why this person, now? A specific answer. "Because we are about to commit to X, and we have not heard from anyone who would carry the cost if X goes wrong."
2. What do you want them to leave knowing? Two sentences max.
3. What do you want to leave knowing? Three concrete questions, written down.
4. What format suits them? A patient advocate, a regulator, a clinician and a community group leader each prefer different formats. A coffee, a structured interview, a workshop, a written response.
5. What changes if they push back? Decide in advance which parts of the plan are negotiable. If nothing is, you are running an information campaign, not an engagement.

A useful conversation is one where at least one of the five answers changes. Otherwise it was a briefing.

Recording is where engagement projects most often go wrong. Two failures are common.

Selective listening. The team writes up the parts of the conversation that confirm what they already thought, and leaves out the parts that do not. The remedy is to make the recording process less reliant on memory: capture quotes verbatim during the conversation, share the notes back to the person you spoke to within a week, and ask them to tell you what you missed or got wrong.

Premature distillation. The team summarises a conversation into a single recommendation before the conversation is even over. The recommendation is then easier to argue with than the underlying evidence. The remedy is to keep the layered structure: what was said, what we think it means, what we propose to do, in three separate columns. Reviewers can challenge each layer separately.

A good engagement record is one a future team member, reading it cold a year later, would describe accurately to a third party. That is the bar.

Engagement is only worth doing if what comes out of it actually changes the work. That last step is the one that fails most often, usually because there is no agreed mechanism for findings to re-enter the project plan.

Three habits that make this work:

• Engagement findings have a named owner in the project team, the same way a risk has a named owner. Without that, no one's job is to act on them.
• Each finding is paired with a decision. Either "we are changing X because of this", or "we are not changing X, for these specific reasons". Both count; a "noted" without either is the failure mode.
• The people you spoke to are told what changed. A short note, a paragraph in a newsletter, a follow-up email. It costs almost nothing and is the only thing that builds trust for the next round.

If a stakeholder you engaged once would speak to you again willingly, the engagement worked. That is the only test that matters.