The process
Six phases. Run in parallel, and revisited on purpose.
Every phase has a question it exists to answer, a deliverable it produces, and a condition that has to be met before the venture advances. Nothing moves forward because time passed.
Diagnose
Is the problem consequential? Market forces, constraints, and structural inefficiency.
Map
Where are the risks and asymmetries? Leverage points and structural advantage.
Architect
What complete venture should exist? Product, capital structure, and operating model.
Build
Can we create a usable proof? Experience, workflows, data model, demonstrable product.
Validate
Will customers care enough to act? Technical, product, and commercial evidence.
Prepare
Should capital be committed? Production scope, GTM, capital plan, investor narrative.
The numbering is nominal. Phases 04–06 routinely send work back into phase 03, because that is what stress-testing an architecture is for. A process that never returns to the design is not testing it — it is decorating it.
Phase 01
Diagnose
What is actually broken, for whom, and what does it cost them today?
We separate the irritation from the inefficiency. Most ideas are a reaction to an annoyance; a venture needs a structural failure — something the current arrangement of an industry cannot fix without changing shape. We size that failure in money, not adjectives, and we look for who is already paying for it in a form they have stopped noticing.
What you get
A written diagnosis of the problem, its cost, the forces holding the broken arrangement in place, and an explicit statement of what would have to be true for it to be worth solving.
The problem is expensive, structural, and currently unowned.
Phase 02
Map
What kills this, and where does a small input produce a disproportionate result?
We map the terrain before committing to a route: incumbents and what they are structurally unable to do, the regulatory and data constraints, the distribution paths that already exist, and the two or three assumptions on which the whole venture rests. Risks get ranked by what they would cost to be wrong about, not by how likely they feel.
What you get
A risk and asymmetry map — ranked kill risks, identified leverage points, and a named first assumption to test.
There is at least one asymmetry we can hold, and the kill risks are testable.
Phase 03
Architect
What is the whole system — not the product, the whole system?
This is the phase the name of the discipline comes from. Strategy, product, technology, unit economics, governance, commercialization and capital plan are designed together, because a decision in any one of them constrains the other five. A pricing model implies an architecture. An architecture implies a cost floor. A cost floor implies who can be sold to. Designing them in sequence is how ventures end up internally contradictory and unfundable.
What you get
The venture architecture: thesis, category, customer, product definition, data model, unit economics, operating model, governance posture, and capital shape — as one coherent document.
The system is internally consistent and the economics survive their own model.
Phase 04
Build
Can someone who has the problem use this and tell us something true?
We build the smallest artifact capable of testing the most consequential assumption — not a demo, and not a product. AI has collapsed the cost of this step, which is exactly why the discipline has to move upstream: when building is cheap, building the wrong thing is the expensive mistake.
What you get
A usable proof of concept: real workflows, a real data model, and enough surface for a practitioner to work in it rather than watch it.
A person with the problem can complete the core task unaided.
Phase 05
Validate
Does anyone change their behavior — and would they pay?
Evidence, gathered in the open. We put the proof in front of operators and buyers and record what they do, not what they say. Enthusiasm is not evidence. A pilot commitment, a signed letter of intent, a purchase order, or a refusal with a stated reason all count; a compliment does not.
What you get
A validation record: what was tested, with whom, what happened, what it falsified, and what it left standing.
The consequential assumption survived contact, or the architecture changed.
Phase 06
Prepare
What does the next stage cost, and is the case strong enough to fund it?
We convert the architecture and the evidence into the package the next stage requires: the production build specification, the go-to-market motion, the capital plan tied to milestones rather than to a round size, and an investor narrative that will survive diligence because it is describing something that already exists.
What you get
Production blueprint, go-to-market plan, capital plan, and investor narrative — plus an honest recommendation on whether to proceed.
Either the case for capital is defensible, or we say so and stop.
The handoff
The Studio's work ends at the proof. Deliberately.
These six phases are pre-proof work: deciding what should exist, designing it as a whole system, and producing enough evidence to justify building it for real. That is a different discipline from making software survive real users, real data and real money — and a studio that pretends otherwise ends up doing both badly.
Everything after the proof — refactoring for production, security and data integrity, governance and IP structure, economic simulation at scale, cap table and control calibration, launch, and investor and acquisition readiness — is BailiwickVibe's 8-Phase Venture Architecture. Same company, same discipline, later stage, and a defined handoff rather than a hazy one.
See the eight phases →Sixty days is the objective. The gates are the guarantee.
We would rather tell you in week two that the problem is not consequential than deliver a beautiful architecture for a business that should not exist.