Method before conclusion
A comparison should begin with the decision a person is trying to make, not the products a publisher wants to list. The method needs to set the audience, criteria, evidence, limits and update conditions before it can responsibly produce a recommendation.
Each future comparison should state its intended user and decision, inclusion and exclusion rules, criteria and weights, evidence sources and dates, any direct testing, limitations, price and regional context, update trigger and commercial disclosure. It should not call one product best when the evidence only supports a narrower conclusion.
The comparison record
- Define the reader, task and decision the comparison is intended to support.
- Publish inclusion and exclusion criteria before deciding which products appear.
- Set criteria that relate to the decision, with red lines where a weakness cannot be traded away.
- Gather current, attributable evidence and label whether it is documentation, direct testing, verified trial, user feedback or unresolved.
- State the result, limitations, commercial context, date checked and conditions that would trigger an update or correction.
Separate fact, observation and judgement
| Type | Example | How it should be shown |
|---|---|---|
| Documented fact | A published support policy or stated compatibility requirement. | Source, date and important condition or region. |
| Direct observation | A stated task completed during a hands-on test. | Method, version, environment and what was not tested. |
| Reader-relevant judgement | An option is easier for a particular stated workflow. | Reasoning, trade-off and audience; not a universal claim. |
| Unknown | Unclear repair route, price, feature limit or update date. | Marked as unresolved, not silently scored as positive. |
Method is part of the answer
Readers should not need to infer how a conclusion was reached. A clear method lets people decide whether the evidence applies to their country, budget, devices, privacy needs and use case. Use the reader guide to assess a product comparison before relying on it.
Method FAQs
Does every comparison need direct testing?
Not always, but a comparison should never imply testing that did not happen. When direct testing is not possible, the page must be clear about the evidence used and the limits of the conclusion.
How often will a comparison be updated?
It depends on what can change the result: product versions, price, availability, support policy, security information, contracts or new evidence. The page should show when relevant information was checked.
Can a comparison include products not sold in every country?
It can, if the page makes availability, price, regulatory, support and regional differences visible. A global label must not conceal a local limitation.
Read comparisons with the right context
Use these guides to understand what a comparison can and cannot tell you. A useful result makes the inclusion, evidence, uncertainty, regional limits and commercial context visible before a reader acts.
