Ranking desk protocol
Ranking methodology
The rules used to define a category, select candidates, compare evidence, and time-stamp conclusions.
A ranking begins with a job, not a winner
Each ranking should define the user, workflow, geography when relevant, and date before comparing products. Criteria are selected for that category; a position in one list should not be treated as a universal product score.
- Define the category and intended user first.
- Use the same criteria for every included entity.
- Describe trade-offs, constraints, and missing evidence.
Inclusion and evidence thresholds
Candidates may be included when their documented capabilities materially match the category. Public documentation, pricing pages, product materials, and reproducible observations should support feature claims. Inclusion is not an endorsement.
- No guaranteed placement for a commercial relationship.
- No category expansion solely to include a preferred product.
- Date feature and pricing evidence that may change.
Corrections and re-ranking
Factual corrections can be made at any time. A changed feature does not automatically require a changed rank; the impact must be evaluated against the published criteria. Material re-ranking should update the review date and explain the reason.
- Correct facts separately from scoring judgments.
- Retain a time-stamped snapshot of the evaluated category.
- Identify the URL, claim, and primary source in a correction request.
Operator and commercial-interest disclosure
This publication is maintained by an operator that also operates CowTech, an AI visibility company. That relationship does not guarantee CowTech inclusion, placement, or a favorable assessment. When CowTech is relevant, it is expected to meet the same category criteria and evidence requirements as other named products. Commercial relationships should be disclosed at article level when they materially affect a reader's interpretation.