Our method
How we test AI tools
Our testing process, what "overview" and "hands-on tested" mean, and how we decide what to recommend.
Two kinds of guide
Every tool guide on Creator Lab carries one of two labels at the top:
| Label | What it means |
|---|---|
| Overview · hands-on test in progress | We've researched the tool: its features, pricing, refund terms, the seller's track record and public user feedback. We have not yet used it on a real project, so we give no rating or verdict. |
| Hands-on tested | We've bought or been given access to the tool and used it on a real project. The guide shows the results, including screenshots or video where possible. |
We never present an overview as if it were a hands-on review.
What a hands-on test involves
- A real job. We use the tool for something we actually need, such as a course explainer, a YouTube thumbnail or a product mockup, not just the vendor's demo template.
- Time to result. We note how long it takes a non-expert to get a usable result.
- Quality check. We compare the result against a free or cheaper alternative where one exists.
- The full price. We go through the real checkout and record every subscription, upgrade and add-on we're offered.
- Support and refunds. Where a tool has a money-back guarantee, we read the terms and note any conditions.
What we look at before recommending anything
- Does it solve a genuine problem for trainers, small businesses or creators?
- Refund and complaint signals. Affiliate networks publish refund rates for some products. A high refund rate is a red flag, and we've excluded products because of it.
- Seller track record. How long the product has existed, how it's supported and whether it's updated.
- Pricing honesty. Aggressive countdown timers and stacks of upsells count against a product.
Corrections
Prices and features change constantly. Every guide shows its last-updated date. If you spot something out of date, tell us and we'll fix it.