# Discovery and briefing

Use this when the idea is vague, the user asks for options, or the brief is not sufficient to lock a direction. Discovery is the shared entry point for web, audiovisual and fixed graphics; it does not restart per medium.

## Goal

Reduce ambiguity with the fewest high-value questions. Do not make the user become a designer. Convert ordinary descriptions into decisions that change the result.

## Read the room first

1. Inspect the conversation and the project's existing brief, direction, tokens, assets and constraints.
2. Extract what is already known; treat existing records as authoritative for their datum.
3. Separate **hard constraints** from **open creative axes**.
4. Ask only about open axes that would materially change the direction, method or rights.

## Purpose and message come first

Establish what kind of piece this is before styling it. A marketing piece, a demo, a tutorial, a story and an editorial piece have different jobs and pacing.

- **Goal**: what the piece should achieve.
- **Audience**: who it is for, what they know, the device and context.
- **Core message / action**: the one thing to remember or do.
- **Deliverable / format(s)**: what is produced and in what format(s). A single brief can cover several deliverables; there is no need to choose one rigid category.
- **Desired perception**: how it should feel.
- **Brand assets / rights**: what may be used, and under which rights.

If the call to action is unclear, a competent execution can still fail the objective. High-value questions when needed: does a graphic get seen **far away or up close**? does a video seek **emotion or process**? does a landing page aim to **clarify or to act**?

## Known vs open idea

- If the idea is **known**, clarify what is invariant, what is preserved and what is interpreted; inspect the real references and materials before proposing changes. Detect internal contradictions before treating a direction as approved.
- If the idea is **open**, the direction phase proposes two or three genuinely distinct options (see [visual direction](visual-direction.md)). Discovery does not need to extract the whole answer first.

Do not invent brand facts, testimonials, metrics or sample data. Hypotheses do not replace facts.

## Brief fields (logical, not a schema)

Collect only what changes decisions:

- subject, audience, primary message/action;
- deliverable(s) and delivery format(s);
- desired perception and ambition;
- existing brand, content and assets;
- technical, accessibility, performance and delivery constraints;
- rights and licensing for any reused or generated asset;
- cost, time and which actions are authorized without asking;
- success criteria: **at least one creative and one technical**;
- declared assumptions.

## Question budget

Use one compact round of only the material unanswered questions, then a second small round only if the answers created new critical uncertainty. There is no minimum count: a sufficient brief needs no interview. If the user delegates the decision, decide and state the assumption. Attach a recommended default to each question so the user can accept them all in one reply.

When the user does not know, do not answer with more abstract questions; offer two or three concrete choices with their implications (for example, restrained vs expressive, precise vs organic, editorial vs product-led).

## Readiness test

A brief is ready when you can answer: what is the subject; who is it for; what is the objective; what must be shown or said; how should it feel; what assets and rights exist; what medium and constraints apply; how will success be judged. A missing answer that changes only a minor implementation detail becomes a declared assumption rather than a blocker.