Ambition with gates.
Every venture here moves through the same six stages, and each stage has a condition that has to be met in the real world before the next one opens.
An idea does not advance because we like it.
Six gates. Each one has a condition that has to be met in the real world before the next one opens — the same questions we ask of our own ideas.
Gate 01 of 06 · Hypothesis
Discover
A problem is written down in one sentence, with the person who has it named.
Fails here whenA thesis nobody can state plainly. If it takes a paragraph, it is not understood yet.
What must become true before it advances
- The problem exists without us. Someone is already paying, in money or in workaround, to avoid it.
- We can describe the buyer specifically enough to go find one this week.
- It does not already belong inside an operating division. If it does, it goes there instead.
Gate 02 of 06 · Evidence
Validate
The thesis is tested against people who would have to pay for it, before anything is built.
Fails here whenEnthusiasm without evidence. Interest is not demand.
What must become true before it advances
- Direct conversations with prospective buyers or operators, not survey data.
- A stated willingness to pay, or a documented reason the current workaround is intolerable.
- A route to market that does not depend on a channel we do not have.
Gate 03 of 06 · Proof
Pilot
A deliberately small version runs in the real world, with a defined end date and a defined kill number.
Fails here whenA pilot with no kill condition. That is not a pilot, it is a commitment in disguise.
What must become true before it advances
- A working thing in front of real users, however unglamorous.
- A metric agreed before launch that decides whether it continues.
- A cost of failure the group can absorb without harming anything else.
Gate 04 of 06 · Structure
Formation
The venture becomes an entity: ownership, IP assignment, operating agreement, capital plan.
Fails here whenStructure built after revenue starts. Retrofitting ownership is the most expensive mistake available.
What must become true before it advances
- A named operator accountable for the outcome.
- IP cleanly assigned, and rights to anything third-party cleared in writing.
- Its own vehicle, so a failure here cannot reach the rest of the group.
- Capital identified and committed, not assumed.
Gate 05 of 06 · Discipline
Operate
It runs as a business with its own numbers, its own reporting line, and its own accountability.
Fails here whenPermanent life support. A venture that only survives on group subsidy is a hobby with an EIN.
What must become true before it advances
- Monthly numbers that a third party could audit.
- A repeatable way of getting a customer, not a series of favours.
- A gross margin that survives paying someone other than the founder to do the work.
Gate 06 of 06 · Leverage
Scale
Capital, systems, and distribution are applied to something already proven to work.
Fails here whenScaling a leak. Volume makes a broken model fail faster, not slower.
What must become true before it advances
- Unit economics that improve, not worsen, with volume.
- Systems and people in place before the demand arrives, not after.
- A clear answer to whether it stays in the group, is capitalised separately, or moves to another division.
Most ideas stop at Validate. That is the system working, not the system failing.
How we build, in fullFour things that do not flex.
-
Its own vehicle
Every venture is held in its own entity. A failure in one cannot reach the others, and anything that outgrows the group can be separated cleanly rather than surgically.
-
A named human on the gate
MIDIOS runs on an internal command layer that handles research, drafting, modelling, and reporting. It does not approve anything. Money, contracts, public claims, and external commitments pass a named person before they move.
-
Rights cleared before build
IP ownership and third-party rights are settled in writing at Formation, not renegotiated once something is working. Retrofitting ownership is the most expensive mistake available to a young company.
-
A kill number, set in advance
Pilots carry an end date and a metric that decides continuation, agreed before launch. An idea that is allowed to run indefinitely has stopped being an experiment.
Seven honest endings.
Not everything that works becomes a company, and that is not a failure mode.
| New company | Formed, capitalised, and operated as a standalone business. |
|---|---|
| Subsidiary | Held under an existing MIDIOS entity where the fit is structural. |
| Joint venture | Built and owned with an external partner who brings something we do not have. |
| SPV | A single asset or project isolated in its own vehicle. |
| Product | Shipped inside an existing division rather than spun out. |
| Partnership or investment | Backed or partnered rather than built, when someone else is better placed to run it. |
| Transfer | Handed to Media Group, Logistics, Technologies, Energy, or Capital, where it belongs. |
If you have something that would survive this.
Tell us what it is and where you think it sits on the ladder. We will tell you honestly which gate it is standing at.