The Strategic Intent Taxonomy

Working draft. Lifted from Valet’s communication-agent skill, where this taxonomy is in production use across executive comms, vendor negotiations, and stakeholder management. Polishing in public.

The problem most communication advice doesn’t notice

Most communication advice operates at the level of grammar and tone. Be clear. Be brief. Be polite. Use the active voice. Bury the lede if it’s bad news, lead with it if it’s good. All true. All necessary. None of it is where the leverage is.

The leverage is one level up. It’s at the intent. Two messages can use identical words and produce opposite results, because the receiver isn’t responding to the words. They’re responding to what they think the sender is trying to do.

Most communication failures I’ve watched, in product reviews, in vendor negotiations, in board prep, in cross-functional escalations, weren’t grammar failures. They were intent failures. The sender knew what they wanted to say. They didn’t know what they wanted the message to do.

So I started naming the moves.

The taxonomy

What follows is the operating set in current use. Nine named intents. Each one has a clean test for whether it’s the right move, a typical structure, and a typical failure mode.

1. Building capital

The intent: investing relational equity for later use, with no immediate ask.

You send a thoughtful note about something the other person did well, an idea they raised in a meeting, a problem you noticed they’re solving. You don’t ask for anything. You don’t bring up the project they owe you. You just leave a deposit.

The test for “is this the right move?” is whether you’d send the same note if you knew you’d never need anything from this person again. If yes, send it. If no, you’re seeding a transaction, which is a different intent.

Typical failure: appending a small ask at the end that ruins the gesture. “Loved your framing in the meeting. By the way, can you take a look at this?” The receiver immediately recognizes that everything before “by the way” was setup. The capital you tried to build evaporates.

2. Creating accountability without assigning

The intent: making a commitment visible so it can’t slip, without saddling a specific person with the weight.

This is the move when something needs to happen, you don’t have authority to assign it, and naming a single owner would create resentment. So you describe the commitment in a way the group has already implicitly agreed to, and let the group sort out who carries it.

Structure: “We agreed [outcome] by [date]. Want to make sure we’re aligned on that before we move on.” No name attached. The accountability is on the group’s prior agreement, not on a person.

Typical failure: sounding passive-aggressive. The fix is to write it as a flat factual recap, not as a question with an edge.

3. Seeding an idea

The intent: introducing a concept slowly so it gets adopted as someone else’s.

You have an idea you think the org should adopt. You also know that if you propose it formally tomorrow, it’ll get reviewed in your name and rejected because the politics aren’t right. So you seed it. You drop a phrase, a reference, a related question, three weeks before anyone needs to decide. You let it incubate.

When the idea comes back as someone else’s, that’s success. You wanted the idea adopted, not credited. Credited is a different intent (see building capital).

Typical failure: seeding too aggressively. If the receiver notices you’re trying to plant something, the idea is now contaminated. The art is making the seed feel like an offhand observation.

4. Protecting a timeline

The intent: surfacing a risk early enough that it can be absorbed without drama.

A timeline is going to slip. You can see it now. The team can’t. Saying nothing means the slip becomes a surprise, and surprises in cross-functional contexts metastasize. Saying it bluntly means triggering a panic that may not be necessary.

The move is to describe the risk and the absorption mechanism in the same breath. “There’s a real possibility we slip on X by 1-2 weeks. If we do, the way we’d absorb it is Y. Want to flag now so it’s not a surprise.” You haven’t predicted disaster. You’ve shown you’ve already thought about the recovery. That changes how the message lands.

Typical failure: framing as “I’m worried we might…” which sounds like you want reassurance. You don’t. You want acknowledgment of the risk and confirmation that the absorption mechanism is acceptable.

5. Lowering the temperature

The intent: de-escalating a heated thread without conceding the substance.

A meeting is getting tense. A Slack thread is sliding. Someone has just said something that isn’t wrong but is too sharp. The intent isn’t to make peace. It’s to slow the room down enough that everyone can hear the substance again.

Structure: acknowledge the underlying concern, restate it in calmer language, surface the decision the group actually has to make. “I think there’s a real tension here between [A] and [B], and it sounds like we’re feeling it. Can we just name the trade-off and decide on it?”

Typical failure: being so balanced that you sound like you don’t have a position. The receiver should know you have a view. They just shouldn’t feel it as heat.

6. Confirming alignment

The intent: verifying agreement on something everyone thinks is decided.

Half the meetings you’ve ever been in ended with “great, we’re aligned” when the participants were aligned on three different things. The confirming-alignment move is sending a written summary back that forces anyone who isn’t actually aligned to surface it now, while the cost is low.

Structure: plain prose, not bullets. Say what you understood. Ask the receivers to flag anything that doesn’t match their understanding. Treat the silences as meaningful.

Typical failure: sending it as a bulleted list. Bullets get scanned. Prose gets read.

7. Surfacing a tension

The intent: naming a real disagreement productively, instead of letting it stay subtextual.

This is the hardest move on the list. You think there’s a genuine disagreement on the table that nobody is naming, because naming it would be uncomfortable. Letting it stay buried is worse.

Structure: name what you think the tension is, name both sides fairly, and ask the room to confirm or correct. “I think we have a real difference of view on [thing]. The case for [A] is [steelmanned]. The case for [B] is [steelmanned]. Where do we actually land?”

Typical failure: failing to steelman the opposing view. If the receiver senses you’ve stacked the framing, the tension stays buried, and now they trust you less.

8. Anchoring expectations

The intent: setting a reference point before a negotiation begins.

You’re entering a conversation where someone’s going to ask you for something, or you’re going to ask for something. Whoever sets the anchor first usually wins. The move is to set the anchor before the formal negotiation starts, in a low-stakes message, so it’s the implicit baseline by the time the real conversation happens.

Structure: mention the reference point as if it’s already obvious. “We typically scope this kind of work over a quarter.” Now “a quarter” is the floor.

Typical failure: anchoring too aggressively. If the receiver feels manipulated, the anchor breaks and you’ve damaged the relationship.

9. Creating optionality

The intent: keeping a path open without committing to it.

You don’t yet know whether you want to do something. Saying yes commits you. Saying no closes the door. The move is to keep the door open in a way that’s honest about the uncertainty.

Structure: “I’d want to do this if [conditions]. Worth exploring whether that’s possible? If yes, here’s how I’d approach it.” You haven’t said yes. You’ve said the conditions under which you’d say yes. That preserves your ability to walk without making the receiver feel rejected.

Typical failure: sounding non-committal. The fix is being explicit about what would make you commit.

Why nine

A taxonomy works only if it’s small enough to keep in working memory and large enough to cover the actual moves people make. Nine fit Miller’s Law plus a small margin for the ones I’d add later. I started with five, the framework gained two more in the first month of production use, then I added the last two over six months. I’ve been at nine for over a year now.

If a tenth move shows up consistently and isn’t a variant of one of these, it gets named. Otherwise the taxonomy stays where it is.

How to use this

Before writing a high-stakes message, name the intent first. Just the name. Then write the message. Re-read it as if you didn’t know what you were trying to do, and check whether it reads as the named intent.

The biggest gain isn’t in the messages where the intent is obvious. It’s in the messages where the writer was unconsciously mixing two intents and confusing the receiver. Naming the intent forces a choice. Choices ship cleaner messages.

---

This is one of several methodology contributions from Valet. The Strategic Intent Taxonomy lives in Valet’s communication-agent skill, where it’s the basis for the register and tone of every executive-grade output the framework produces.