How We Test Lookup Services

No standardized hands-on benchmark has been completed for this launch. This page defines the protocol a future test must satisfy and prevents source review from being mislabeled as product testing.

Current test status

We reviewed public documentation and access surfaces. We did not submit a balanced set of addresses, measure match accuracy, time queries, purchase subscriptions or test cancellation and support.

Future test dataset

A valid benchmark will use authorized addresses with known ground truth, separated into professional domains, personal webmail, aliases, changed jobs and deliberately absent records. Raw addresses will not appear in public artifacts.

Measures

The protocol will distinguish lookup success, correct identity, correct field, stale field, false match, no result, time to result and total checkout cost. A single “accuracy” percentage will not hide these outcomes.

Equal treatment

Every compared provider must receive the same permitted inputs, browser conditions, geography, date window and scoring rules. Paywalls and account requirements are recorded as outcomes rather than bypassed.

Publication threshold

We will publish a test-backed ranking only when all providers have complete comparable coverage, the methodology and date are visible, and another reviewer can reproduce the calculations from redacted artifacts.

Sources used for this page

Provider statements describe what the provider says its product does. They are not converted into independent test results.

  1. Google high-quality review guidance

    Calls for evidence, expertise, measurements where available, pros and cons and useful comparisons. Checked 2026-09-03.

  2. Google people-first content guidance

    Editorial reference for original, useful and transparent content. Checked 2026-09-03.