Conversation Planner

Updated October 23, 2026.

Fill in the form to see your plan.

How it works

The planner asks for nine things and puts them in one block of text:

1. Who and what situation. A plan for “a conversation” is too loose to test, so name the person and the moment.
2. The one thing to say differently. One change is easier to carry into a real conversation than five.
3. The decision point. The part of the message where the other person decides how to read the rest: the opening, the ask, the reason, the timing, the way out or the close.
4. The failure if rushed. Confusion (they cannot tell what you want), defensiveness (they hear blame) or misread intent (they read your motive wrongly). Naming one tells you what to watch for.
5. Before and after sentences. Written out in full, so you can read both and see what changed.
6. The disclosure line. One sentence beginning “What I am doing and why:”. It applies the briefing’s test: you should be able to say it to the person. If you leave it blank, the tool composes one from your decision point and failure, and tells you it did.
7. What you will look for. A behavior someone could see or hear within the time window. The briefing’s standard examples are in the list: they paraphrase your point back accurately, they ask a follow-up question instead of restating their position, their reply gets longer, the thread ends in a decision, you finish the sentence you started.
8. The observability check. If your own wording contains one of these words, the tool warns you: feel, felt, trust, respect, engage and empower (with endings such as feels, trusted, respected, engagement and empowered), comfortable, rapport and buy-in. These describe inner states, which you cannot watch. The check only flags the listed words. It cannot confirm that a sentence is observable.
9. The time window. Two hours, 24 hours or 48 hours. The briefing’s standard is a check within 24 hours.

The Copy button is added by the page script. If copying is blocked in your browser, the plan sits in a text block you can select and copy yourself.

Why write sentences down. Keysar and Henly (2002) had speakers say ambiguous sentences, such as one that could be read two ways, to listeners and estimate how often they were understood. Speakers expected to be understood more often than they were, while overhearers who knew the speaker’s intention did not overestimate in the same way. The study used spoken ambiguous sentences in a laboratory. It does not test planning, rewriting or any tool like this one, so whether a written plan improves understanding is Not reported. The planner uses the finding as a reason to check what the listener did instead of assuming they understood.

Nothing in the planner is a threshold from research. The word list and the options are this tool’s own rules of thumb, not research findings or benchmarks.

Worked example

The boxes above hold an illustrative plan. A team member needs more time on a report and tends to bury the request in background. The change is to put the request first. The decision point is the request, and the risk if rushed is defensiveness, because a vague opener can sound like an excuse.

With “The thread ends in a decision rather than another round” selected, the planner finds nothing to flag: the line is one of the listed observable examples. Had the check read “Priya feels respected”, the planner would warn that “feels” and “respected” describe an inner state and suggest something visible, such as a reply that gives a date. The people and details are invented for this example.