Methodology
How the Build Score works
A weighted editorial rating out of 100. It is our judgement, applied consistently — not a product rating, a guarantee or investment advice.
| Dimension | What it measures | Weight |
|---|---|---|
| Market | Size, durability and local availability of demand. | 15 |
| Economics | Contribution margin after real delivery costs — not revenue. | 15 |
| Accessibility | How much capital, licensing and skill stands between you and day one. | 10 |
| Recurring revenue | Share of revenue that renews without a new sale. | 10 |
| Customer acquisition | How hard and how expensive it is to win the first and the fiftieth customer. | 10 |
| Operations | Delivery complexity, staffing load and failure modes. | 10 |
| Scalability | Whether the second, tenth and hundredth unit get easier or harder. | 10 |
| AI leverage | How much of the work AI can genuinely absorb today. | 10 |
| Defensibility | What stops the next operator from copying you. | 5 |
| Risk | Regulatory, concentration, seasonality and liability exposure. | 5 |
| Total | 100 | |
Sourcing standards
- Every published number is either cited, or labelled as a modeled estimate.
- Government and industry-association data is preferred over vendor marketing.
- Prices are research estimates for a defined model account — never a quote.
- Demonstration companies used in builds are fictional and labelled as such.
Verdict scale
- BUILD — the economics and the path to customers both hold up.
- BUILD WITH CONDITIONS — it works, but only if specific conditions are met.
- SKIP — the numbers or the acquisition math don't survive scrutiny.
Educational content with modeled assumptions. Costs, taxes, labor, insurance, legal and licensing requirements vary by location — verify for your jurisdiction.