The Market Transfer Framework

Most venture risk is demand risk. The Market Transfer Framework removes it by starting from models that already work somewhere — then doing the hard part: deciding where they travel, what must be rebuilt, and when to stop.

The thesis: arbitrage on proof

Every new venture answers two questions. Does anyone want this? Can we deliver it profitably? The first question kills most ventures, and it is the one a transfer thesis answers in advance. When a model has paying users, visible retention, and explainable economics in one market, the demand question is settled there. What remains is a different and more tractable problem: whether the conditions that made it work exist — or can be built — in a market the incumbents are not serving well.

This is not cloning. A clone copies the surface and hopes the mechanics came along. A transfer decodes the mechanics, checks which ones survive the move, and rebuilds the rest deliberately. The difference sounds subtle. In practice it is the difference between a venture and a screenshot.

Inside Exiid the framework sits in the Opportunity Engine. Its output is not a build plan — it is a scored thesis that either earns a validation sprint or goes back on the shelf. Everything downstream — Validation, Product, AI, Growth, Scale — runs faster because this stage refuses to pass along ambiguity.

Part one: selection

Selection runs on five criteria. The first four are gates — binary, and a miss on any one stops the thesis. The fifth is a dial that sets priority among theses that clear the gates.

| # | Criterion | The kill question | | --- | --- | --- | | 1 | Model legibility | Can you explain the machine, not the vibe? | | 2 | Structural gap | Will the gap survive the first competitor launch? | | 3 | Transfer distance | How many layers must be rebuilt, not translated? | | 4 | Unfair access | What can you do in week one that a funded cold entrant cannot? | | 5 | Timing window | Why is this possible now and not three years ago? |

1. Model legibility

A transferable model has visible machinery: the offer, the buyer, the pricing motion, the acquisition loop, the retention behavior, and the moment where value compounds. If the reference business can only be described as a story, it cannot be decoded — and what cannot be decoded cannot be transferred.

2. Structural gap

Underserved does not mean empty. It means demand already exists — people pay with money, time, or workaround effort — while the category standard stays fragmented, offline, overpriced, slow, or untrusted. The filter is durability: a gap that exists only because nobody has launched yet closes the day someone does. A structural gap persists because incumbents cannot or will not close it.

3. Transfer distance

Two markets can want the same outcome and still sit far apart in payment behavior, logistics, trust defaults, regulation, and acquisition channels. Distance is not a reason to stop. It is a cost to price in. Short-distance transfers ship fast and compete on execution. Long-distance transfers take longer and earn deeper moats, because every rebuilt layer is a layer the next entrant must also rebuild.

4. Unfair access

Distribution, domain authority, regulatory fluency, supply relationships, capital with patience — any of these can count. Enthusiasm does not. The advantage must change acquisition, trust, speed, or execution quality in measurable terms, and it must be assignable to a named partner with explicit incentives before anything gets built.

5. Timing window

Strong transfers ride an unlock: payment rails maturing, smartphone penetration crossing a threshold, a regulation changing, an incumbent stumbling, AI collapsing a cost line that made the model uneconomic at local price points. If nothing changed recently, the honest question is why the gap is still open — and the honest answer is often that it is not a gap.

Part two: the Localization Stack

The most expensive mistake in model transfer is localizing the wrong layer first. Teams translate the interface — the cheapest, most visible layer — and leave trust and distribution, where transfers actually die, on origin-market defaults.

The stack orders the work. Lower layers travel; upper layers get rebuilt.

| Layer | What it holds | Treatment | | --- | --- | --- | | L0 — Core mechanic | The behavior loop that creates value | Travels intact. If it needs changing, the thesis is wrong. | | L1 — Unit economics | CAC, basket size, margin, payback | Recalculated from local inputs. Never imported. | | L2 — Trust layer | Payment methods, guarantees, social proof, brand signals | Rebuilt from zero. Trust does not travel. | | L3 — Distribution | Channels, partnerships, acquisition loops | Rebuilt around local channel reality. | | L4 — Surface | Language, content, design idiom | Translated last. Cheapest layer, least decisive. |

Two rules govern the stack. First, work bottom-up: confirm the mechanic, recalculate the economics, then rebuild trust and distribution before touching the surface. Second, budget honestly: L2 and L3 routinely consume 70 percent of localization effort in long-distance transfers. A plan that allocates most of its effort to L4 is a translation project wearing a venture costume.

Part three: failure modes

Five patterns account for most dead transfers. Each has a tell.

The Photocopy. The team copies features instead of mechanics. Tell: the roadmap mirrors the reference product's interface while nobody can explain why its retention loop works. Corrective: decode before you build — if the loop cannot be named, the transfer has not started.

The Phantom Gap. The gap exists only because nobody launched yet. Tell: the pitch leans on "first mover" rather than a structural reason incumbents stay out. Corrective: name what blocks a well-funded fast follower. If nothing does, the gap is a countdown.

The Imported P&L. Origin-market assumptions — CAC, basket size, payment completion, return rates — pasted into the new market. Tell: a financial model with suspiciously familiar numbers. Corrective: rebuild L1 from local data or local tests, never from the reference deck.

The Translation Trap. L4 polish substituting for L2 work. Tell: a beautifully localized interface bolted onto checkout, guarantees, and proof signals designed for a market that already trusted the category. Corrective: spend on trust mechanics first — cash on delivery, local guarantees, recognizable proof — and let the surface lag.

The Enthusiasm Moat. The partner advantage turns out to be excitement. Tell: access claims that cannot be scheduled — no named introductions, no committed channel, no signable asset. Corrective: convert the advantage into dated commitments before the build starts, or rescore criterion four at its true value.

Part four: the Transfer Scorecard

Theses that clear the gates get scored before any validation spend. Six dimensions, each rated 1 to 5.

| Dimension | 1 looks like | 5 looks like | | --- | --- | --- | | Model legibility | A story about a company | A mapped loop with named economics | | Gap durability | Open because untried | Open for structural reasons incumbents cannot fix | | Transfer distance | Four-plus layers rebuilt blind | One or two layers, with local data in hand | | Unfair access | Enthusiasm and a network "somewhere" | Dated, signable, channel-specific commitments | | Timing | No recent unlock | A named unlock under 24 months old | | Proof point clarity | "Revenue, eventually" | One measurable behavior, with a kill threshold attached |

Read the total in three bands. A score of 24 or higher earns a validation sprint. Between 18 and 23, fix the weakest dimension before spending — a stronger pitch does not repair a weak score. Under 18, shelve it. One veto rule overrides the math: any single dimension at 1 stops the thesis regardless of total, because transfers fail at their weakest layer, not their average.

What the framework refuses to do

The scorecard does not predict success. It prices risk before capital and attention get committed, and it makes the no-go cheap. A thesis that scores well still has to survive contact with the market — that is validation's job, and it starts with the smallest proof point that demonstrates the transferred behavior, never with a roadmap.

Models keep. They sit on the shelf without decaying, waiting for the layer that blocked them to shift. The discipline is letting them sit until it does.

Read next

• Is this model worth transferring? The four-gate checklist that feeds this framework's selection stage.
• RECON before roadmap. What happens after a thesis clears the scorecard.
• Ethical transfer versus cloning. Where decoding ends and copying begins.