Research
Context
AI is compressing the cost of building software and improving operations. Those advantages matter, but they are becoming easier to reproduce. Customer access, trust, workflow position, and accumulated market history do not compress in the same way. They are built over time, through repeated participation in a market.
This changes what an investor can underwrite. A distribution network is not only a channel for selling a product. At sufficient density, it becomes a panel: a recurring view into pricing, demand, timing, risk, and behaviour across a sector. The network produces both market access and proprietary intelligence.
We believe this is the beginning of a distinct investment category. The opportunity lies in businesses whose distribution compounds into infrastructure — particularly multiplier businesses that sit between fragmented operators and the markets they serve.
Research
Software once captured value because it was scarce and expensive to build. AI is changing that equation. Products can now be reproduced faster, and operational improvements diffuse more quickly across a market. Distribution remains slower to assemble because it depends on trust, repeated use, and accumulated relationships.
We study businesses where those relationships are not merely an acquisition channel but the principal asset: networks whose strategic value increases as products change, and whose market position survives the replacement of any single tool.
A shifting hierarchy of defensibility
| Layer | Investment characteristic |
|---|---|
| Product | Fast to build; increasingly easy to reproduce |
| Operations | Valuable execution layer; improvements diffuse |
| Distribution | Trust and workflow position accumulate over time |
| Telemetry | Market meaning improves with density and history |
| Infrastructure | The network becomes a reference point for the market |
Some businesses reach one customer at a time. Others serve customers who each represent an additional network: tax advisers, brokers, managed service providers, property managers, accountants, and other intermediaries. One relationship can create visibility into dozens or hundreds of downstream businesses.
We call these multiplier businesses. We are mapping where their distribution is fragmented, which workflows earn recurring participation, and when their position can support a broader platform without weakening the trust that made the network valuable.
A list of customers is not yet an intelligence asset. The transition occurs when a network crosses a density threshold: participation becomes broad and frequent enough to reveal changes in pricing, demand, timing, and behaviour across a market.
At that point, distribution begins to function as a panel. Its history matters as much as its scale. A network recording a market today accumulates context that a later entrant cannot recreate simply by spending more money.
How a distribution asset matures
| Stage | What compounds |
|---|---|
| Access | Reach into a defined customer group |
| Engagement | Evidence of recurring workflow use |
| Density | Representative coverage of a market |
| History | A calibrated record of change over time |
| Reference | A trusted benchmark others rely upon |
What we avoid
What we seek
Distribution produces information at several levels. Engagement data can establish whether a network is active. Practice-level signals can reveal operational changes, demand, or succession. Aggregated over a representative panel, those signals can describe the state and direction of a market.
The investment question is not who has the most data. It is who has permissioned access to meaningful, recurring observations — and enough calibration history to interpret them responsibly.
A hierarchy of information value
| Layer | Signal | Use | Duration |
|---|---|---|---|
| Engagement | Usage | Diligence | Short |
| Practice | Operational change | Selection | Medium |
| Panel | Market movement | Prediction | Compounding |
| Reference | Shared benchmark | Infrastructure | Durable |
The underwriting discipline is to distinguish information exhaust from information advantage. Only the latter deserves to be valued as an asset.
Falling build costs create a period in which products can be assembled quickly but durable networks still cannot. This makes distribution-rich businesses easier to misprice: they may be valued as ordinary software or services while their deeper asset is the market position they have accumulated.
We are developing the frameworks required to recognise that position early, help operators deepen it, and underwrite the transition from useful product to trusted network to market infrastructure.