Reproducible by design
Research method and disclosures
How questions are selected, sources are sampled, test cases are prepared, observations are recorded and limitations are reported.
1. Define a falsifiable question
Each study starts with a specific claim, comparison or observable behaviour. The scope is written before the result is interpreted.
2. Record the tested conditions
The sample, products, versions, dates, locations, accounts and relevant settings are documented. A result describes those conditions, not every possible user or query.
3. Preserve evidence
Where safe and permitted, observations are retained as source links, exports, screenshots, test cases or downloadable data. Dynamic results are time-stamped.
4. Separate observation from interpretation
The reported result states what happened in the tested sample. Explanations, hypotheses and recommendations are identified as interpretation.
5. Report limitations and conflicts
Material commercial relationships, incomplete samples, unstable systems and plausible alternative explanations belong in the study, not in a footnote added later.
6. Update without rewriting history
Meaningful corrections and reruns should state what changed and when. Older observations may remain useful evidence of how a system behaved at a specific time.
TomSEOTools does not claim that a schema type, score, format or writing technique guarantees rankings or AI citations. Visibility alone does not prove authority, accuracy or impact.
Important limitation
Written and tested by Tom Petryk
Tom Petryk researches how search engines and generative systems evaluate, retrieve and cite web information. He has more than 20 years of experience in digital marketing and web development and publishes the tools, methods and test data used in TomSEOTools studies.