Skip to main content
AI Health Guide
Menu

Editorial Independence

The most-asked question we get from clinicians and patients alike: does the affiliate money distort what you recommend? This page is the long-form answer.

What we earn money from

We earn affiliate commissions when readers sign up for a product through a tracked link on this site. Commission rates differ by vendor, by program, and by the action the reader takes (free signup, paid signup, qualifying first visit). Our typical range across health-vertical affiliate programs is $5–$120 per qualifying signup; some clinical-software vendors do not have affiliate programs at all.

We do not charge vendors to be reviewed. We do not charge for inclusion in any "Best X" guide. We do not accept payment for editorial placements. We do not accept free product in exchange for coverage. If a vendor offers free access to facilitate a hands-on review, we accept only when the access is offered to legitimate journalists at large, not on a quid-pro-quo basis, and we disclose it inline on the review.

How rankings are decided

Rankings are produced by a scoring rubric documented on our methodology page. The rubric covers clinical evidence, regulatory status (FDA, FTC, state telehealth), pricing transparency, EHR integration depth (for B2B clinical tools), and hands-on testing experience. Affiliate-commission amount is not a rubric input.

In practice this means: a vendor with no affiliate program can rank #1, and frequently does. Several of our top-ranked B2B clinical scribes (Abridge, Nuance DAX, Suki) sell through enterprise procurement and have no consumer affiliate program at all. They are top-ranked because of their KLAS scores, deployment footprint, and clinician interviews — not because we earn anything from them.

Where affiliate amounts vary widely between similarly-ranked tools, we publish the variance. See for example our best AI mental-health apps ranking, where BetterHelp's higher CPA does not move it above Wysa on the clinical-evidence axis.

The firewall

  • Vendors do not preview drafts. Reviewed companies do not see copy before publish and do not edit copy after publish. The exception is factual corrections — if a vendor flags a price error, regulatory date error, or factual mistake, we'll correct it and note the correction inline.
  • Affiliate teams do not influence editorial. We do not have a sales team. The editor (Stephan Kulik) sets affiliate program selection, but ranking is produced independently by the rubric.
  • Negative reviews ship as readily as positive ones. Our heightened mental-health affiliate disclosure surfaces FTC settlements (BetterHelp 2023, Cerebral 2024, Talkspace 2024) inline on every review of those companies. We don't suppress the bad news to protect the affiliate revenue.
  • We name what we don't know. Where we haven't tested a product hands-on, the review says so. Where the clinical evidence is thin, the review says so. Where pricing changes weekly, the review says so.

What happens if we get this wrong

If we publish a recommendation later shown to be biased by commercial relationship, we apply the procedure in our corrections policy: re-evaluation against the methodology rubric, inline correction notice at the top of the affected page with the original publish date, and (where the error materially affected ranking) revised rank order with explanation. The correction stays inline on the page in perpetuity — we do not silently fix history.

Who decides what stays in this policy

This page is signed off by Stephan Kulik in his role as editor of AI Health Guide. Material changes require a corresponding entry in the corrections log. Soft revisions (clarifying language, adding examples, updating commission ranges) are recorded by `dateModified` on this page only.

Where to verify

Questions

If you're a clinician, patient, vendor, or regulator and have a question about a specific recommendation on this site, the fastest path is [email protected]. We acknowledge inside 48 hours. If you believe a published recommendation is materially biased by commercial relationship, please cite the page URL and the specific claim — we'll re-run it against the rubric and publish the outcome.

Featured: BetterHelp Get Matched ↗