Methodology
How we research
Every page on CloudedSite tells you what we know about a tool and how we know it.
Two kinds of evidence, always labelled
- Used first-hand. J. Bond has used the tool in his own work. The page includes his screenshots and notes, dated.
- Research-based. The page is built from the vendor’s documentation, pricing pages and independent sources, all listed at the end. It says so at the top.
Many pages mix both. When they do, each first-hand observation is marked as one.
What we check
- Price and plan limits, from the vendor’s own pricing page, with the date checked.
- What the tool does well, for a specific kind of user and task.
- Where it falls short, including limits, reliability problems and learning curve.
- Data and privacy terms that matter for client or business data.
- Alternatives, so you can see what else fits.
How verdicts are written
No AI tool is best for everyone. Each verdict names who a tool is for and who should skip it. We avoid star ratings unless a score is based on first-hand use; research-based scores are labelled as the editor’s judgement.
Keeping pages current
AI tools change quickly. The date at the top of each page shows when it was last checked. We re-check the most-read pages regularly and update the date when we do.
Money and independence
Some links earn a commission if you buy through them, at no extra cost to you. That never changes a verdict, and a tool can be recommended without an affiliate program. Details are in our affiliate disclosure.
Corrections
If something is wrong or out of date, please tell us. We correct factual errors and note significant changes on the page.