Yardstick Research tear-sheet / healthcare RCM cohort

Methodology · how we score · rubric weights in plain sight · vendors received this sheet seven days before publication and could flag factual errors, never rankings

Phreesia

Identity

• Name: Phreesia, Inc. (NYSE: PHR)
• Founded: 2005 by Chaim Indig and Evan Roberts; IPO July 2019
• HQ: Wilmington, DE (fully remote workforce)
• Funding: Public company. FY26 revenue $480.6M; first full year of positive GAAP net income ($2.3M); Adjusted EBITDA $101.5M. Acquired AccessOne (Sept 2025, $160M cash) for healthcare-receivables financing.
• Employees: 1,789 full-time as of January 31, 2026 (per 10-K)
• Leadership: Chaim Indig (CEO + co-founder), Evan Roberts (President, Provider Solutions + co-founder), Balaji Gandhi (CFO), David Linetsky (President, Network Solutions)
• Archetype: Public-company patient-intake heritage platform extending into RCM-adjacent automation (eligibility + benefits, time-of-service collections, AccessOne-financed payment plans) and conversational AI (VoiceAI for call management). Not a pure-play RCM vendor; RCM is the monetization layer on top of a 20-year patient-intake business.

Total score: 65 / 100

(75 raw weighted - 10 pricing-transparency hard penalty)

• Stage fit:
• Foundation (<40 readiness): no - Sales-led, vendor-managed implementation; multi-year auto-renewing contracts; quote-only pricing. Foundation-stage buyers cannot self-evaluate or self-onboard.
• Pilot (40-59): conditional - yes if pilot is a single product surface (e.g., eligibility verification only) at a mid-market practice ($25M+ revenue) with an existing EHR Phreesia natively integrates with; no for AI-first pilots evaluating VoiceAI maturity.
• Scale (60-79): yes - this is the sweet spot. Mid-market and enterprise ambulatory groups and health systems wanting a single front-end intake-through-payment platform with the broadest published EHR connector library in the cohort.
• Optimization (80+): yes with caveats - strong scale and integration breadth match optimization-stage needs; the AI surface is administrative-workflow AI, not clinical-documentation AI, so optimization buyers wanting AI-native architectural depth need to look at AKASA, CodaMetrix, or Tennr instead.
• One-line verdict: The cohort's heavyweight incumbent - broadest EHR connector library, largest scale by far (~180M patient visits / year), HITRUST r2 since 2017 - weakened by post-2022 AI bolt-on architecture, hard pricing-transparency penalty, and the May 2026 securities class actions hanging over the public-company parent.

Headline numbers

Dimension scores

Pricing detail

• Public price anchor: None. /pricing is a quote-request form. Copy reads "Customizable packages for healthcare organizations of all sizes" with "Request Phreesia pricing" form and "REQUEST A DEMO" CTA.
• Third-party-reported model (not vendor-confirmed):
• Starting band: $250-$300/month per third-party aggregators (Capterra-derived, ITQlick, SelectHub).
• Common configurations: $300-$800+/month depending on modules, integrations, locations.
• Annual baseline (10-user practice): $3,600-$9,600 per ITQlick estimate.
• Reported contract terms (Capterra-aggregated): Annual contracts with 60-day cancellation window before renewal; auto-renewal after initial term; pre-transaction fees on optional modules (texting, reminders, payments); add-on hidden costs.
• No free tier and no free trial. Engagement starts with sales-led demo.
• Wayback regression check: May-2024 snapshot (referenced in existing healthcare-rcm scoring) shows identical "Customizable packages... Request Phreesia pricing" framing - baseline opaque, no regression. Phreesia has been quote-only for at least 2 years.

Integrations

• EHR / PM (16 named bidirectional connectors): Epic, Oracle Health (formerly Cerner), Athenahealth, NextGen, eClinicalWorks, MEDITECH, ModMed, Greenway Health, AdvancedMD, Veradigm Connect Developer, Office Practicum, eMDs, Compulink, Azalea Health, Unlimited Systems, Flatiron.
• Open standards: HL7, FHIR, CCD, CSV "and more" for non-listed systems.
• Epic marketplace: App Orchard listing since 2019; integration now surfaced via Connection Hub within Epic Showroom (App Orchard / App Market are legacy names).
• Payer connectivity: 1,000+ payers for E&B verification (primary, secondary, specialty benefits).
• Notable EHR gaps (no native connector): Allscripts (separate from Veradigm), Kareo / Tebra, Practice Fusion, drchrono, CharmHealth, ChartLogic, iSalus, OncoEMR. Integration for these requires Phreesia services-team work or customer IT effort via open standards.
• Depth-vs-breadth caveat: [THIRD-PARTY] athenaOne competitive note describes Phreesia integration in some scenarios as "PDF data push" rather than fully structured bidirectional. Depth varies by EHR. Right-of-reply item.
• Cohort-peer integrations: None. Phreesia does not list any healthcare-rcm cohort peer (AKASA, Cohere Health, Adonis, Tennr, CodaMetrix, Infinitus, Hippocratic AI, Fathom, Nym, etc.) as a native integration. The integration value is to the customer's EHR, not to other RCM vendors. integrates_with: [] is correct per the rubric.

Editorial assessment

Phreesia is the heavyweight incumbent of patient intake, and its extension into RCM-adjacent automation makes it a legitimate but specific candidate for the healthcare-rcm cohort. The strongest case is at a $250M+ ambulatory health system or large MSO that wants a single front-end platform spanning scheduling through patient payment, with the broadest published EHR connector library in the cohort (16+ named bidirectional integrations plus Epic Showroom listing) and a HITRUST r2 + SOC 2 Type 2 + PCI DSS L1 posture that has held continuously through three full audit cycles since 2017. The 4,658 customer organizations and ~180M annual patient visits are not marketing inflation - they're SEC-disclosed quarter-over-quarter and constitute the largest scale claim in the cohort by a wide margin.

The bounded weak spots matter for AI-priority buyers. First, the AI surface is bolt-on, not architecture. VoiceAI shipped September 2025 (nine months mature at this writing); foundation model undisclosed; no published independent evaluation; no latency or accuracy benchmarks; no hallucination guardrail documentation. ai_native_score=25 places Phreesia at the cohort floor on this axis. The cohort's AI-native candidates (Tennr, Hippocratic AI, CodaMetrix, Infinitus, Nym, Fathom, Adonis, AKASA's GenAI re-platform) all ship as AI-first products and are founded post-2018. Phreesia is the legacy intake platform adding AI; that is a different bet. Second, the AI scope is administrative-workflow AI, not clinical-documentation AI - buyers comparing VoiceAI to ambient-listening SOAP-note products (Abridge, Suki, Ambience, DeepScribe, DAX Copilot) are comparing different categories. Third, the May 2026 securities class actions and the cut FY27 guidance are not a customer-operations problem today, but a multi-year RCM contract is a multi-year bet on the vendor; public-company parent under earnings + investor-litigation pressure is a risk worth pricing. Fourth, the pricing opacity is hard-penalty material: three independent third parties cite a $250-$800+/month band, zero of those numbers come from Phreesia, and Phreesia's own copy invites SMB buyers who cannot self-size.

Relative to the cohort: Phreesia is in a different weight class than the pure-play GenAI RCM peers. AKASA is the strongest enterprise GenAI RCM vendor for Epic / Cerner shops with Cleveland Clinic-scale rollouts. CodaMetrix is the autonomous-coding specialist at academic medical centers. Cohere Health is payor-side prior authorization. Adonis is the SMB-friendlier AI-first RCM platform. Tennr is DME / specialty / infusion intake. Phreesia is the only vendor in the cohort that started as a patient-intake platform and ships front-end RCM as an extension; it competes on integration breadth, scale, and balance-sheet stability rather than on AI architectural depth. For a buyer who weights workflow_integration_depth (25%) and data_compliance_posture (20%) heavily and AI capability (15%) less, Phreesia scores well on the weighted total - the 4/4 on integration depth and 3/4 on compliance posture carry a lot of weight. For a buyer who weights AI capability heavily, the 2/4 on ai_capability_depth drags the score noticeably.

Revisit if: (1) Phreesia discloses VoiceAI foundation-model architecture and PHI-in-training opt-out default, (2) the FY27 guidance situation resolves with clarity (Network Solutions stabilizes or is divested), (3) the May 2026 securities class actions reach early dismissal or settlement, (4) a competitor (Tebra, athenahealth) ships an equivalent intake-plus-RCM bundle at price-transparent terms, or (5) Phreesia publishes a per-visit or per-provider price anchor on /pricing.

Right-of-reply gaps

The following [UNKNOWN] items should be sent to Phreesia during the factcheck pipeline:

• VoiceAI foundation model and architecture - Which LLM / model provider(s) power VoiceAI? Is it OpenAI, Anthropic, Google, AWS Bedrock, custom? Model-version-management posture for clinical-triage routing?
• VoiceAI accuracy / latency benchmarks - Published intent-recognition accuracy, average call-handling time, escalation-to-human rates beyond marketing case studies?
• VoiceAI languages supported - "Multiple languages" is unspecified; which languages and comparative accuracy by language?
• PHI training-data opt-out default - Is customer PHI excluded from VoiceAI / AI-eligibility model training in the standard BAA by default?
• Hallucination guardrails for VoiceAI - Guardrails on AI output affecting clinical-triage routing or billing-affecting outputs? Audit log for AI-generated decisions?
• Pen-test report publication cadence - Last pen test date? Reports available under NDA?
• Data residency - Where is PHI processed and stored? Multi-region options?
• EHR integration depth per system - Which of the 16 listed EHR connectors are bidirectional with structured-data write-back vs. one-way push vs. PDF / document-style? (Per the athenaOne competitive note, depth appears to vary.)
• Per-visit, per-provider, or per-AHSC pricing anchor - Will Phreesia publish even a "starting at $X per provider per month" anchor on /pricing?
• Auto-renewal terms - Is the 60-day cancellation window reported by Capterra reviewers accurate? Negotiable?
• VoiceAI rollout scale - How many customers have VoiceAI live in production? Patient-call volume across those deployments?
• Network Solutions risk to customer-facing operations - Operational firewall between Network Solutions (under investor-litigation pressure) and intake / RCM customer-facing product lines? Will FY27 guidance pressure affect product roadmap velocity, account-management staffing, or support quality?