Our Methodology

Editorial Independence

Some platforms in our reviews have affiliate programs that pay us commissions. Upheal pays $300/referral, SimplePractice pays $100+, Freed pays $50, Alma pays $500, Wysa pays 15%, and Mentalyc offers revenue sharing. We disclose every affiliate relationship on every page where affiliate links appear.

Our commitment: Affiliate commissions never influence rankings. We rank platforms by clinical evidence, user experience, pricing value, and safety protocols. If the best tool has no affiliate program, it still ranks first. If the highest-paying affiliate has weaknesses, we report them.

Example: Alma pays $500/referral — the highest in our coverage. Yet we report its directory saturation problems, unreliable billing, and slow customer support. Blueprint, which has no affiliate program, earns positive coverage for its free EHR and low-cost AI notes.

Evaluation Criteria — Consumer Apps (B2C)

We assess consumer AI therapy apps across six dimensions:

1. Clinical evidence (30% weight): Published peer-reviewed studies, RCT quality, effect sizes, sample sizes, journal prestige. A tool with 30+ papers (Wysa) scores higher than one with internal data only.
2. Therapy modality coverage (20%): Range and depth of evidence-based approaches (CBT, DBT, ACT, etc.). Does the app implement genuine therapeutic techniques or just branded chat?
3. Safety protocols (20%): Crisis detection capability, escalation procedures, responsible messaging about limitations. Apps that lock crisis features behind paywalls are penalized.
4. User experience (15%): App store ratings, user reviews, conversation quality, onboarding ease, engagement metrics where available.
5. Pricing transparency (10%): Clear pricing, reasonable free tier, value for premium features, no deceptive billing practices.
6. Regulatory status (5%): FDA, CE-mark, NHS endorsement, or other regulatory recognition.

Evaluation Criteria — Therapist Tools (B2B)

1. Note quality and accuracy (25%): How well does the AI generate clinically appropriate documentation? Does it capture therapeutic nuance or produce generic notes?
2. HIPAA compliance and security (25%): BAA availability, SOC 2 certification, data handling practices, encryption standards.
3. Template and modality support (15%): Variety of note templates (SOAP, DAP, BIRP, custom), support for diverse therapeutic modalities.
4. Workflow integration (15%): EHR compatibility, recording methods, ease of use during sessions, Chrome extension reliability.
5. Pricing value (10%): Cost per note, free tier availability, total cost of ownership including add-ons and fees.
6. Unique features (10%): Session analytics (Upheal), measurement-based care (Blueprint), training capabilities (Lyssn), alliance insights (Mentalyc).

Clinical Evidence Evaluation

We rate clinical evidence on a four-tier scale:

• Gold: 5+ RCTs with control groups in peer-reviewed journals
• Silver: 1-4 peer-reviewed studies or large observational data
• Bronze: Internal data or validated instruments used (no independent platform studies)
• Unrated: No published clinical evidence

See the full Clinical Evidence Scorecard for every platform's rating.

YMYL Standards

Mental health is a YMYL (Your Money or Your Life) topic. We apply heightened content standards:

• Crisis resources (988 Suicide & Crisis Lifeline, Crisis Text Line) appear on every page
• Mental health disclaimer appears above the fold on every comparison and review page
• Affiliate disclosure appears on every monetized page
• We never claim AI tools "treat" or "cure" mental health conditions
• We explicitly state limitations and appropriate use cases for every tool reviewed
• Clinical claims are attributed to specific published studies

What We Don't Do

• We don't accept payment for reviews or rankings
• We don't allow platforms to pre-approve our content
• We don't suppress negative findings to protect affiliate relationships
• We don't recommend tools for conditions beyond their validated use cases
• We don't provide medical advice or therapeutic recommendations

Update Schedule

All reviews and comparisons are reviewed quarterly. The next scheduled update is July 2026. If a significant event occurs (platform shutdown, major research publication, pricing change, safety incident), we update immediately regardless of schedule.

Contact

Questions about our methodology? Corrections? Contact us at hello@bestaitherapy.ai.