Write the Same Result Three Different Ways

You run an internal survey and share what came back. Distribution chart, observations, a summary line. It makes a clean single page and it goes out to everyone. People read it. Nothing happens.

This is written for people responsible for AI adoption inside an organization. It is also for anyone on the receiving end of a share like that who could not work out what they were supposed to do with it. The premise is that the reader is not the problem.

One line, three different available actions

Say the result is "almost nobody here is playing the checking role." One fact. But the people who can act on it sit at three levels, and what each can actually do is entirely different.

An executive can allocate a person and hours to that role. A manager can decide which member of their own team to ask. Someone on the floor can either take up the checking themselves or name the person they want checking their output.

Send one page to everyone and those three arrive tangled together. So everyone reads it as somebody else's job. The executive assumes the floor will handle it, the manager waits for a directive, the floor assumes leadership decides. No bad intent anywhere. The page simply had no addressee.

So the report splits into three

The org reports we deliver to client organizations are written for three audiences separately: leadership, managers, and the people doing the work. The underlying numbers are identical. What a reader can do tomorrow is not, so the prose is not either.

Splitting it has a side effect worth having. When a conclusion refuses to yield three different actions, that conclusion was never going to produce any action at all. Trying to write the three versions is itself a test of whether the observation was worth writing.

"Please drive adoption" never reaches a calendar

There is a second thing the real deliverable does. Every action in the 90-day plan carries the role who owns it and the time it takes.

On one occasion it came out like this. One item at fifteen minutes for a Shadow Adopter. One at two hours a week for a Systemizer. One at four hours a month for a Sponsor. Those figures are design allocations rather than measured records. They still do work.

"Please drive AI adoption" lands on nobody's calendar, because with no duration attached there is no answer to when. Fifteen minutes fits into today. Two hours a week becomes something to raise with your manager. Four hours a month becomes an agenda item for the leadership meeting.

Naming the time also lets the person say no. That is a feature. A refusal tells you the action does not fit the current load, which you learn far sooner than you would from three months of no reply.

Who to ask is not settled by job title

Splitting into three audiences still leaves the real question open: within "for managers," who do you ask, and for what? Fill that in from the org chart and you will usually be wrong.

RELACTIS is the assessment we built for that. Twenty-four questions, about four minutes, and you get seven factors (Creation, Crossover, Autonomy, Advocacy, Systemization, Verification, Engagement) plus which of fifteen types you sit closest to. It costs nothing and needs no signup. Take it alone to see your own shape, or run it across a team for the distribution by role.

The three time budgets above suit three particular seats. The assessment narrows the candidates; the last step, choosing who to ask, stays with the person asking.

One thing for tomorrow

Take one survey result you already shared and write it three times, one line each. One line for leadership, one for managers, one for the floor. Each line has to be something the reader can do tomorrow.

If all three come out as the same line, the exercise did not fail. The conclusion has not reached an action yet. Knowing that is worth the rewrite on its own.