Content Classification
Hands-on Reviews
A page is labeled as a hands-on review only when the editorial team has directly accessed the product and completed documented test tasks.
A hands-on review should identify, where applicable:
- Test date
- Plan or account type used
- Device, operating system, or browser
- Tasks completed
- Relevant screenshots
- Results observed
- Limitations encountered
- Pricing check date
Research-based Overviews
When direct testing has not been completed, the content must be labeled as a research-based overview. These pages may summarize official product information and practical use cases, but they must not imply first-hand experience.
Comparison Guides
Comparison pages must state whether they are based on hands-on testing, official product information, or a combination of both. Products should be compared using the same criteria wherever possible.
Evaluation Criteria
- Intended use case
- Ease of use
- Core workflow
- Output or feature quality
- Integrations
- Pricing and plan limitations
- Suitability for individuals or teams
- Strengths
- Limitations
- Alternatives
- Review effort required
Pricing
Pricing must include a “Last checked” date. We do not guarantee future pricing. Readers should check the product provider for current terms.
Guided Recommendations
The guided finder evaluates confirmed task, audience, ease-of-use and optional must-have requirements against our dated catalog. Missing task or audience tags remain unknown rather than proving a mismatch. When a specific task guide is selected, its editorial candidates appear first within each eligibility group, followed by documented task coverage, your reuse or ease preference, and confirmed-condition coverage. Alphabetical order breaks ties. Affiliate relationships do not affect eligibility or order. This is an explainable research order, not an output-quality score.
Numeric budgets and workloads require plan-specific evidence. An over-budget finding for a selected plan does not exclude all plans from that provider. Existing software does not automatically mean that additional usage is free. Unsupported or out-of-date capability claims cannot satisfy a must-have condition.
Task guides, practice materials and tests
Our six task guides include fictional exercises, illustrative editorial references, reusable prompts and human checkpoints. Reference answers are teaching materials, not captured provider output. Workflows use manual handoffs unless an integration is separately verified. Completing a checklist does not certify the output as correct.
A published task test must record the input and permission to use it, tool, plan, visible version, date, settings, output, generation time, correction effort, reviewer and limitations. Compare tools serving the same stage under comparable conditions; an editing assistant is not a complete substitute for a drafting or recording workflow. No task-level performance winner is declared without completed evidence.
Cost worksheets use values entered by the reader. They separate additional monthly spending from annual seat prepayment and clearly identify excluded taxes, exchange rates and overages.
A recorded free tier does not prove that a particular task or export is free. We flag unverified plan coverage separately. Decision evidence older than 90 days and pricing evidence older than 30 days require another check for strict matching. These are editorial review intervals, not guarantees that a provider has made no changes.
The current finder uses structured guidance and keyword suggestions, not generative AI. Changing requirements requires an explicit user action. No match percentage or hands-on testing is implied.
Affiliate Relationships
Affiliate relationships must be disclosed clearly. They do not permit fabricated experience, guaranteed outcomes, or misleading product claims.
Corrections and Updates
Substantive updates should update the dateModified field. We do not change an article's update date merely to make it appear recent.