Build three
worlds at once.
Decide later.
A workspace for strategists building parallel brand directions — archetype, expression, activation, in loops, with co-pilots and a pressure-tester. This prototype covers loop 1: foundation, three worlds, the archetype co-pilot.
"Three to four worlds in parallel. Each one earns the right to exist or doesn't. The system makes the difference visible."
The brand before its expression.
Everything three worlds will share. Fill in what you have; you can refine before locking.
Untitled
Project-level instruments shared across all three worlds. Build them in any order — later instruments may require earlier ones.
Each world resolves the consumer insight in its own way. Run the archetype co-pilot on any world to begin.
Archetype co-pilot
A quick brief, then we begin.
The co-pilot needs a minimum context to be useful. Tell it about the brand; the conversation does the rest. Nothing here saves to your project.
Competitive set assistant
First, the
competitive set.
Positioning is triangulation, and the triangle's third point is who you're positioning against. The positioning tool requires a structured competitive set as input — it grounds the work, surfaces the jobs-to-be-done, and gives the pyramid's Frame of Reference and Point of Difference real substance.
What's missing
No competitive set has been built for this project yet.
When the competitive set tool finishes, you'll be returned here to continue.
A quick brief, then we begin.
The positioning co-pilot needs context plus a competitive set. Provide both; the conversation does the rest. Nothing here saves to your project.
Positioning assistant
Surfaced from the competitive set — particularly the status quo and indirect entries. These are jobs the consumer is already hiring some product or behavior to do today. The positioning pyramid's functional benefits are this brand's case to be hired for these jobs.
A quick brief, then we begin.
The archetype co-pilot needs minimum context to be useful. Provide what you have; the conversation surfaces the rest. Nothing here saves to your project.
How this world feels
Why this pair, what it isn't
First, archetype
and positioning.
Voice is downstream of archetype and positioning — it's the verbal expression of both. Without them committed, voice work is unmoored. The voice direction tool requires both as input.
What's missing
—
Once prerequisites are committed, return here to continue.
A quick brief, then we begin.
Voice direction is the lightweight, foundational pass on how a brand speaks — attributes, principles, and do/don't pairs. A richer verbal identity system comes later as its own engagement.
Voice direction assistant
How this brand sounds
How this brand speaks
In practice
Strategy first,
then names.
Naming without strategy produces names that sound nice and mean nothing. The naming tool requires archetype and positioning as input; voice direction is strongly recommended. The territories the co-pilot proposes are direct extensions of the brand's strategic foundation — without that foundation, the territories are arbitrary.
What's missing
—
Once prerequisites are committed, return here to continue.
A quick brief, then we begin.
Naming work needs strategic context plus a tactical brief. Required fields ground the work in strategy; optional brief fields sharpen the output. Trademark and domain verification happen externally.
Each field below sharpens the co-pilot's territory framing and pressure-testing. Provide the ones that matter for your engagement; skip the rest.
These inputs shape the creative gravity field the name lives within. Especially valuable for collaborative naming sessions — they let participants self-correct rather than relying on a strategist to police every contribution. Fill in the ones your engagement deserves.
Naming assistant
Why this name
Finalists considered
Verified and unverified
Territory framing
Foundation input contract
The workspace is the synthesis hub in a hub-and-spoke architecture. Standalone tools — research synthesis, JTBD validation, social listening analysis, competitive scanning — produce structured outputs that the workspace accepts as inputs to strategic foundation work. This page defines the shape those outputs take.
The shape of a foundation input
Every foundation field in the workspace can carry three things:
- Content — the actual text or structured data (consumer target paragraph, brand idea sentence, competitive set object, positioning pyramid, etc.).
- State — one of
hypothesis,partial,validated. Reflects what evidence has populated this field. - Evidence — an array of attribution entries naming what produced this content.
A field with hypothesis state is supported by pre-validation evidence (competitive scan, social listening, strategist inference). A partial field has mixed evidence — some validating, some not. A validated field has been confirmed by validating evidence (qualitative interviews, quant validation, stakeholder confirmation). Fields without an entry default to validated — the legacy assumption that committed content is trusted.
Evidence attribution shape
Each evidence entry has the following structure:
{
"source_type": "qual_interview" | "qual_video_diary" | "quant_validation"
| "stakeholder_interview" | "competitive_analysis"
| "social_listening" | "research_synthesis"
| "strategist_inference",
"source_label": "Short human-readable identifier",
"note": "Optional free text expanding on the evidence",
"confidence_value": 67, // optional, percentage for quant validation
"added_at": "2026-05-15T14:30:00Z"
}
source_type determines the evidence's role in state computation.
Pre-validation sources (competitive_analysis, social_listening, strategist_inference)
support hypothesis state. Validating sources (qual_interview, qual_video_diary,
quant_validation, stakeholder_interview, research_synthesis) support partial or
validated state depending on volume and convergence.
The state model
Producing inputs the workspace accepts
Any tool that wants to feed the workspace's foundation produces output in the following form, keyed by foundation field name:
{
"consumer_target": {
"content": "Wellness-curious adults, 28-42, urban...",
"state": "hypothesis",
"evidence": [
{
"source_type": "social_listening",
"source_label": "Cluster: 'tired of performing wellness'",
"note": "47 mentions across 8 communities, sentiment +0.6",
"confidence_value": null,
"added_at": "2026-05-12T10:15:00Z"
}
]
},
"consumer_insight": {
"content": "I want energy without the crash...",
"state": "partial",
"evidence": [
{ "source_type": "qual_interview", "source_label": "Interview 03, Maria, 34, Brooklyn", "note": "Direct quote", "added_at": "2026-05-13T..." },
{ "source_type": "qual_interview", "source_label": "Interview 07, Daniel, 38, Oakland", "note": "Paraphrased; same tension", "added_at": "2026-05-13T..." }
]
}
}
Valid foundation field names: consumer_target,
consumer_insight, positioning,
brand_idea, portfolio_scope,
growth_horizon, constraints,
competitive_set_raw, research_inputs.
Structured artifacts (competitive_set,
positioning_pyramid, voice_direction,
naming) have their own internal schemas — see the co-pilot
outputs for those.
Design principles
Tolerant. The workspace accepts what it recognizes from an input and ignores fields it doesn't. Extra fields don't break the import. Missing optional fields are filled with defaults.
Attributed. Any field populated from outside the workspace carries evidence attribution. The strategist can always trace a claim back to what supported it.
Promotable. Fields are not locked at hypothesis state. As validating evidence arrives, the strategist promotes them — either manually (through the workspace UI) or via further imports that add validating evidence.
Honest. The state of every field is visible in the project brief and other shareable artifacts. Stakeholders see what's hypothesis and what's validated — they're invited to pressure-test the hypotheses rather than approve a falsely polished foundation.
How this world resolves the project's consumer insight. The thing that makes this world genuinely different from its siblings.