Email Finder Leaderboard

How the score is built

Each metric is normalized across the field to a score from 0 to 100. Price and speed are inverted since lower is better, and speed uses a log scale so async waterfalls and instant APIs stay comparable. The weighted sum of the five metrics gives the ColdIQ Score. Accuracy carries the most weight: a cheap email that bounces still costs you sender reputation.

Marketplace providers are tested live. ColdIQ live test, October 2026: 400 real B2B contacts (200 EU / 200 US across 6 verticals and every seniority), every provider called through its dedicated email endpoint on the ColdIQ marketplace, two passes 72 hours apart with the median kept. We measure per-region find rate, median lookup latency, and agreement between providers as an accuracy signal, and run every returned email through two independent verifiers. Providers we can't call ourselves are compiled from published benchmarks (independent find-rate and bounce-rate tests, with the sample size on each source) and official pricing pages, with sources on every row of the dataset.

Measures differ by origin. Accuracy for live-tested rows is the share of a provider's emails that another independent provider also returned. Published rows use the valid or non-bouncing share reported by the cited source, so the two are close but not identical signals. Prices are each provider's own self-serve list price per found email, except LinkUp, which publishes no per-email price: its row uses the ColdIQ marketplace rate.

Scores carry real uncertainty. For live-tested rows the coverage margin is the actual Wilson 95% interval of the run (its exact hit counts show on hover); for compiled rows it comes from published benchmarks. Propagated through the weighted aggregate, that puts a margin of roughly 14 to 18 points around every score. Treat providers within a few points of each other as statistical ties: the order is our best estimate, not a proven gap.

Figures remain estimates: coverage swings with your ICP, geography and list quality, and email data shifts weekly. Every pricing figure is re-checked against the provider's official page - each row's sources carry their verification date (latest 2026-10-12) - and rows are rescored as new tests land.

The fastest way to know what works on your market: run the same lookup against several providers side by side.

Work at one of these providers? If a figure looks wrong, dispute it - point us at the source (or provision a test API key) and we re-run the protocol and publish whatever it says, within 7 days.