Methodology - How We Research and Rank

Data first, prose second

Every bot entry is a structured fact set: pricing model and current plans, custody model (who holds your API keys), documented security incidents, feature flags (grid, DCA, signals, futures, backtesting, paper trading), exchange support, regional availability and sources with dates. Reviews are AI-drafted from that fact set only - the model may not invent numbers, plans or history. Drafts are fact-checked against the database before publication, and every fact row carries a verification date.

Where facts come from

Provider pricing pages (fetched and dated, not cached), provider documentation and terms, community threads (Reddit via public mirrors - heavily spam-filtered), published incident reporting, and operator research. Quality is flagged per bot: verified, reported or unverified. Unverified facts are labeled in reviews, never presented as confirmed.

Scoring

Editorial scores (0-10) are weighted: safety 25%, pricing value 18%, features 17%, ease of use 12%, exchange coverage 10%, transparency 10%, support 8%. Safety covers custody model, key handling options (IP whitelisting, OAuth), and incident history including disclosure quality. Bot categories are scored within their own lane - a free exchange-native grid bot does not "lose" to a $130/mo platform on features it never claims. User ratings are shown separately, smoothed with a statistical prior, and never mixed into editorial scores.

Freshness

Prices and plans change frequently. We run daily reachability checks on every listed domain, re-fetch pricing pages on a rolling cycle, and regenerate reviews whenever facts change. Every review shows its version, generation date and fact-verification date.

Independence

Commercial relationships never influence scores, ranks or scam listings. Providers can submit factual corrections with sources via contact; we verify and correct within 7 days. We do not remove scam or dead listings in exchange for anything.