About AI Tried It
AI Tried It is an independent site that assesses AI tools the way a small business owner would actually use them. We started it for a simple reason: the AI tool review space is broken. Search for almost any AI tool and you’ll find the same pattern — long “best of” lists written by people who never opened the product, rewriting the vendor’s marketing page and collecting a commission either way.
We do the opposite. We research each tool in depth — official documentation, pricing, independent reviews — and publish what we find, including the parts the marketing page leaves out. Where we can test hands-on (free tiers, trials, public demos), we do. If a tool is genuinely good for a small business, we say so. If it’s overhyped, we say that too.
Our testing standard
This is the bar every guide on this site is measured against. We’re a new publication: our first guides are deep research assessments, and each article carries a “How we tested” box that tells you exactly what was done for that guide — what we verified, and what we couldn’t. Full hands-on account testing is the standard every guide works toward.
1. We start from the official source
Marketing pages are optimistic by design. Before we publish, we read the vendor’s official documentation and pricing page so that limits, usage caps, and real monthly costs are accurate as of the test date. Where pricing isn’t public, we say so instead of guessing.
2. We test what’s testable
We try the free tier, the free trial, or the public demo wherever we can access them — the closest thing to the experience you would get. We don’t use press accounts, vendor walkthroughs, or sponsored demos, because those don’t reflect what a paying customer sees. When account testing isn’t possible for a guide, the “How we tested” box says so plainly.
3. We cross-check against independent reviews
We never rely on the vendor’s word alone. Claims about features, pricing, and limits are checked against independent reviews and documentation — and when sources disagree, we tell you they disagree instead of picking the convenient one.
4. We never invent results
This is the rule everything else rests on. If we couldn’t test a feature — because it needed an enterprise plan, a region we don’t operate in, or an integration we don’t have — we say so explicitly. We never describe outcomes we didn’t observe.
5. We re-test when tools change
AI tools change fast. When a tool we’ve covered gets a major update or pricing change, we re-test the parts that changed and update the verdict with a new “last updated” date.
What we don’t do
- No paid placements. Vendors cannot pay to be tested, to be ranked higher, or to change a verdict.
- No listicles of tools we haven’t researched. Every tool on our site is assessed against official sources — and each guide’s “How we tested” box says exactly what we did.
- No screenshots presented as ours that aren’t. Figures awaiting our own test captures are clearly marked as placeholders.
- No stale verdicts. Every page carries a visible “last updated” date, and we correct mistakes publicly when readers point them out.
How we’re funded
Testing tools properly takes time, and time costs money. We’re funded two ways: display advertising, and affiliate commissions on some of the links on this site — meaning if you buy a tool through one of our links, the vendor may pay us a commission at no extra cost to you.
Here’s what that does not buy: a better verdict, a higher ranking, or a place on the site at all. Our testing process is the same for tools that pay us and tools that don’t, and you can read the full details in our affiliate disclosure.
Why trust us
Trust isn’t claimed; it’s demonstrated. Three things to check us on:
- Evidence over adjectives. Our guides show their work — what we tested, how, and what the limits were.
- Honesty about limits. We tell you what we couldn’t test and where our experience might not match yours.
- Corrections in the open. If we get something wrong, we fix the page and note the correction. You can report an error any time via our contact page.