How we test
"Deep tested" means four things, on every page.
1. The same tasks on every tool
Each category has a fixed task list. Every tool in that category gets exactly the same tasks, in a fresh account, on the plan a beginner would pick (usually the free plan or the vendor's free trial). If a task needs a plan we didn't use, we say "not tested" instead of guessing.
2. Timed and recorded
We screen-record the whole test with a timer on screen. Times and click counts come from the recording. Short clips from it sit next to each result so you can see what happened.
3. Scored with a published rubric
Scores come from weighted categories we publish in advance. Raw times sit next to every score, so you can reach your own conclusion if you weigh things differently. Email marketing rubric.
4. Dated and re-tested
- Last tested and prices checked dates are shown at the top of every review.
- We re-check prices and plans regularly, and re-run the full test every 6 months or after a major product change.
- Every check goes in the public test log, and each page has a "What changed" section.
What we don't do
- No vendor pays for a review, a score or a position. Affiliate commissions are explained on our disclosure page.
- We don't copy vendor feature lists as findings. Facts about prices and plans are checked on the vendor's official pages and linked with the date we checked.
- We don't send test emails to anyone except inboxes we own.
Honest limits
A short trial with a small test list can't prove long-term deliverability, uptime or how support behaves at scale. We say so on every review.
Who tests, and how we use AI
Tests are run and written by the DeepTested test desk. We publish under the brand rather than personal bylines. We don't invent reviewer names or photos.
AI helps us with narration (video voiceover is AI-generated), first drafts and editing. It does not produce results: every time, score, price and limit comes from our own recording or from the vendor's official page, and is checked before publishing.