Methodology / Evidence first

Show the work.

Every recommendation should make its evidence, limits, incentives, and update history visible.

Last updated August 3, 2026

01

Research starts with operators

We begin in the places practitioners describe the work before it gets cleaned up for marketing. That includes Reddit, specialist communities, public workflow repositories, talks, interviews, and detailed practitioner posts.

Community conversations help us find recurring problems, workarounds, objections, and dissent. They are qualitative evidence, not settled fact.

02

Claims are verified at the source

Product behavior, API details, pricing, policies, and technical limits are checked against current official documentation and hands-on tests when access is available. Role and market claims are compared with active job descriptions, original datasets, and named practitioner experience.

Clay is an important primary source for the category it helped define. It is not the only authority on what GTM engineering has become.

03

Every important claim has a type

  • Fact: directly supported by credible evidence
  • Community signal: a recurring qualitative practitioner observation
  • Inference: a conclusion drawn from several inputs
  • Editorial judgment: the recommendation engineerGTM is prepared to defend

04

Tool rankings follow a fixed rubric

We evaluate workflow fit, reliability, implementation effort, extensibility, total value, support, and governance. The rubric is defined before affiliate economics are considered.

A tool may be best for one workflow and wrong for another. We prefer recommendations by use case over a fake universal winner.

05

Flagship work must clear a publication gate

Our internal score covers reader usefulness, evidence, technical accuracy, original analysis, practical artifacts, clarity, counterarguments, trust, and freshness. Flagship pieces must score at least 90 out of 100 and pass every factual hard gate.

Publishing cadence never justifies lowering the standard.

06

Updates and corrections

Fast-changing tool and AI articles are reviewed more frequently than evergreen operating principles. Meaningful changes should be recorded on the page. If you find an error, contact hello@engineergtm.com with the page and supporting evidence.

Commercial relationships are governed by the disclosure policy.