# How we review company profiles

_Updated 2026-10-18_

We compare company profiles with available licensing-channel criteria using deterministic rules rather than AI-generated guesses. Criteria record their source and review status where evidence is available; unverified criteria remain provisional. Unknown answers stay unknown. Results are labeled "preliminary fit," never a probability. A person reviews every profile before any introduction.

## Matching dimensions

Industry, team size definitions, operating history, retained-record history, geography, systems, record types, language, problem-to-outcome depth, and rights and authority.

## Types of evidence

| Type | Meaning | Effect |
| --- | --- | --- |
| Explicit requirement | A source clearly states a criterion. | Can determine whether stated requirements are met. |
| Preference | A preferred size, region, or sector. | Influences fit, never a hard cutoff. |
| Fit signal | A descriptive opportunity signal. | Informative only. |
| Unknown | Not published. | No positive points; never an automatic pass. |

## Uncertainty and freshness

- Unknown answers reduce evidence coverage rather than counting as a pass.
- We never infer U.S. employees from headquarters, records history from company age, or English from country.
- Criteria older than our freshness window are re-verified before a strong label is shown.
- Vague published wording is retained as text and flagged for confirmation.

## Human review and rights

Rights and authority are evaluated separately from commercial fit. A reported lack of rights blocks any introduction. Our team reviews every profile, and overrides are recorded with a reason.

## Referral compensation disclosure

Some channels publish referral programs. We may receive compensation if an introduction leads to an agreement, and we disclose any applicable relationship before an introduction. Ranking ties are broken by name, never by potential compensation.

## Frequently asked questions

### Is a high score a probability of a sale?

No. A preliminary fit score is not a probability of acceptance or payment.

### Do you publish your scoring configuration?

No. We describe the method here; specific internal configurations and leads are private.

## Sources

- [NIST, AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework) (reference reviewed 2026-10-18; not buyer eligibility verification)

## Related

- [What qualifies](/what-qualifies)
- [How it works](/how-it-works)
- [About](/about)