Research methodology

How we research and structure small business ideas.

How we use AI-assisted discovery, structured attributes, confidence labels, and quality controls to maintain the SmallBizIdeas.top library.

Last updated: July 23, 2026

AI-assisted discovery, clearly labeled

AI is used to discover candidate business models, identify common variants, group similar ideas, and convert unstructured research into a consistent schema. AI makes broad coverage possible; it does not make every generated claim true.

One idea must represent a distinct operating model

We do not count simple wording changes as new ideas. A variant should differ meaningfully in customer, delivery method, capital requirements, regulation, skills, or economics before it becomes its own entry.

What every entry contains

Current confidence status: directional

The launch library is designed for comparison and discovery. Costs, licensing, competition, and market demand must be checked against current primary sources in the user's location before action.

How the library should improve

Future research cycles will add source-level citations, geographic availability, conflicting-source records, confidence scoring, expired-data alerts, and category-specific review rules. Low-confidence records should be visible for browsing but should not dominate top matches.

Editorial independence

Inclusion and matching are not purchased. Future affiliate relationships or advertisements must be clearly labeled and must not change the match score.