Hub Vendor Selection Checklist: How Pharma Patient Services Leaders Evaluate Platforms in 2026

Hub vendor selection is the single most consequential decision a launch team makes for a specialty therapy. The pharma hub and patient access support services market was valued at roughly $3.6 billion in 2025 and is projected to grow at a 10.8% compound annual rate through 2035 [https://market.us/report/pharma-hub-and-patient-access-support-service-market/]. The vendors competing for those dollars now run the gamut from full-service operators handling 600+ programs to API-first technology platforms to AI workforce solutions that sit underneath whichever model a manufacturer chooses.

Yet most manufacturers still run RFPs against a generic capability matrix. That approach produces unusable responses, because hub vendors no longer fit a single template. The difference between a strong hub and a weak one shows up in time-to-therapy distributions, prior authorization workflow depth, and what data you can take with you when the contract ends — not in the marketing slides.

At Neon Health, we sit on the operations side of that table. Our AI workforce runs benefit verification, prior authorization, financial assistance enrollment, and onboarding workflows inside hub programs every day, which is the vantage point this checklist comes from. What follows is the framework we use when patient services leaders ask us how to evaluate platforms — vendor-neutral, 10 domains, RFP-ready.

What Does a Hub Vendor Actually Do? Scope Before Selection

A modern patient services hub functions as the single point of contact for a patient starting a specialty therapy. The core service stack includes intake and enrollment, benefit verification (BV), prior authorization (PA) support, copay and patient assistance program (PAP) administration, adherence and patient education, REMS compliance, nursing or clinical support, specialty pharmacy triage, and data reporting back to the manufacturer [https://www.grandviewresearch.com/industry-analysis/pharma-hub-patient-access-support-service-market-report].

How a hub assembles those components defines its operational model. There are three dominant patterns:

The decision underneath the operational model is what to insource versus outsource. Financial assistance is almost always outsourced because it requires firewalled administration. Education and clinical support are often kept in-house. Operational workflows like BV and PA submission used to default to outsourcing — but AI workforce platforms have made internal operation a realistic option in 2026 [https://www.pharmaceuticalcommerce.com/view/patient-support-hub-services-continue-to-evolve].

Pick the model before you write the RFP. Manufacturers who skip this step receive responses that are impossible to compare, because vendors are quoting different scopes.

The 10-Domain Hub Vendor Evaluation Checklist

Every domain below includes a short framing, what to ask, what good looks like, and a red flag. Score each vendor on a 1-5 scale per domain. Weight the domains by what your launch actually needs — for a rare disease product, therapeutic fit and clinical support dominate; for a high-volume specialty therapy, time-to-therapy and PA workflow depth carry more weight.

1. Therapeutic Fit and Launch Experience

Hub services are not interchangeable across therapeutic areas. A vendor with 30 oncology launches has built different muscle than one with 30 dermatology launches. Payer mix, prescriber behavior, and patient journey complexity all change.

Ask: How many launches in this therapeutic area in the last five years? Peak patient volume per program? What was the typical buy-and-bill versus specialty pharmacy mix? Will you share the program list under NDA?

What good looks like: Documented launch experience in directly adjacent indications, named program references, and a willingness to put case study data behind an NDA.

Red flag: Generic case studies that swap the therapeutic area name without changing the operational details.

2. Time-to-Therapy Performance

This is the metric every other metric serves. Specialty pharmacies report turnaround times of 2 to 3 days for a clean prescription and approximately 5 days when intervention (PA, clinical coordination) is required [https://www.jmcp.org/doi/abs/10.18553/jmcp.2022.28.11.1244]. Integrated health system specialty pharmacies hit a median 12 days from order to therapy initiation, while external transfers run 18 days — a six-day, statistically significant gap [https://www.jmcp.org/doi/10.18553/jmcp.2024.30.4.352]. Hub-coordinated programs vary widely depending on workflow integration.

Ask: What is your median time from referral to first fill across active programs? What is the P95? What does the distribution look like for prescriptions that require PA versus those that don't? Will you share TAT data for a comparable launch?

What good looks like: Vendors who share both medians and P95s, and who segment by PA-required versus clean Rx. The variance matters more than the average — a hub with a 7-day median and a 28-day P95 has a workflow problem the average hides.

Red flag: Vendor reports only an average and refuses to share the distribution.

3. Benefit Verification Depth and Speed

Real-time eBV is now table stakes for high-volume programs. But the technology label hides meaningful differences in coverage breadth and therapy-specific intelligence. EDI 270/271 transactions return basic eligibility but miss step therapy sequences, quantity limits, and PA requirements specific to the drug.

Ask: What share of BVs complete in real time versus requiring manual outreach? How many payers do you cover real-time? Do you return therapy-specific rules (step therapy, quantity limits, site of care restrictions) or only generic eligibility?

What good looks like: Real-time eBV across the major commercial and Medicare/Medicaid payers, with documented therapy-specific rule capture and a defined SLA for manual outreach on payers outside the real-time network.

Red flag: "We use EDI 270/271" with no therapy-specific layer on top.

4. Prior Authorization Submission and Tracking

Prior authorization is the workflow most directly responsible for therapy delays. Surescripts data shows prescribers handle 39 PAs per week, consuming 13 hours of staff time, with 88% of pharmacists reporting that PA frequently or consistently delays treatment [https://surescripts.com/prior-authorization-challenges-data]. Manufacturers carry the consequence in abandonment and time-to-therapy metrics, even though the work happens in the prescriber's office and the hub.

Ask: What share of PA submissions are ePA (electronic prior authorization) versus fax or portal? What's the PA approval rate at first submission? At appeal? What's the median PA decision turnaround by payer category?

What good looks like: ePA-first workflow with documented portal automation for payers that don't accept ePA, named approval-rate benchmarks (with payer-mix context), and a defined appeal pathway with success-rate data.

Red flag: Heavy fax and manual workflow disguised as "white-glove service." Manual PA volume is a leading indicator of long tails in time-to-therapy.

5. Financial Assistance and Copay Program Administration

Out-of-pocket cost is the most common cause of specialty drug abandonment after PA delays [https://www.jmcp.org/doi/10.18553/jmcp.2023.29.5.449]. Hub vendors run the copay programs, foundation enrollment, and free drug eligibility workflows that determine whether a patient with a $4,000 deductible actually starts therapy.

Ask: Do you administer copay cards in real time at the pharmacy? How do you handle foundation enrollment when a foundation has a waitlist? What's the annual reverification workflow at year-end? How do you handle eligibility changes mid-year?

What good looks like: Automated re-enrollment workflows, integrated foundation search across all major specialty foundations, real-time financial eligibility checks, and a defined year-end reverification protocol that doesn't create an abandonment cliff in January.

Red flag: Manual annual reverification with no proactive outreach.

6. Technology Architecture and Integration

The platform question used to be a binary — proprietary system or Salesforce-based [https://www.pharmaceuticalcommerce.com/view/patient-support-hub-services-continue-to-evolve]. In 2026 there is a third option: AI workforce platforms that layer onto your existing CRM and operate as integrated staff rather than as a separate portal.

The integration surface that matters: EHR connectivity (for prescriber-facing flows), payer portals, specialty pharmacy systems, foundation databases, and the manufacturer's CRM or data warehouse.

Ask: What integration types are supported (REST APIs, HL7, FHIR, SFTP)? Can you share API documentation pre-contract? What is your data warehouse strategy? Where does our data live?

What good looks like: Documented APIs available pre-contract, SOC 2 Type II certification, named integration partners (EHR vendors, specialty pharmacy systems), and a clear data ownership model.

Red flag: "Our portal" with no API access. All data flowing through the vendor's interface means you cannot integrate, cannot extract, and cannot replace.

7. AI and Automation Maturity

By early 2026, AI tools moved from pilot programs to operational technology across most patient services hubs [https://market.us/report/pharma-hub-and-patient-access-support-service-market/]. The marketing language hasn't kept up — every vendor claims AI capabilities now, so the question shifted from "do you use AI" to "where in the workflow is AI actually doing the work, and what does the human-in-the-loop design look like?"

Ask: Name specific workflows where AI executes end-to-end (not just suggests or summarizes). What's the escalation rate to human staff? What's the accuracy benchmark on outbound payer calls or portal navigation? How do you handle exceptions?

What good looks like: Specific workflows named (outbound payer call for BV, portal navigation for PA submission, form completion from enrollment fax), accuracy metrics with denominators, and a documented exception protocol.

Red flag: "AI-powered" marketing language without workflow-level specificity, no accuracy metrics, no escalation rate.

This is where Neon Health's AI workforce model differs from traditional automation. Rather than bolt AI dashboards onto a manual workflow, AI workers handle end-to-end tasks — making the payer call, navigating the IVR tree, extracting the benefit detail, updating the system of record — with humans handling only the exceptions. The right question is not whether a vendor has AI, but whether AI is replacing labor or just augmenting it.

8. Compliance, Firewalls, and Pharmacy Affiliation

Several major hub vendors are owned by or affiliated with specialty pharmacies. That creates inducement and steering concerns that HHS evaluates on a case-by-case basis [https://www.pharmaceuticalcommerce.com/view/patient-support-hub-services-continue-to-evolve]. Specialty pharmacy-affiliated hubs are not disqualifying, but the firewall documentation is non-negotiable.

Ask: Disclose all pharmacy ownership and corporate affiliations. Provide firewall documentation if applicable. What are your HIPAA, HITRUST, and SOC 2 certifications? What are the BAA terms? What is the audit cadence and who runs it?

What good looks like: HIPAA, HITRUST, and SOC 2 Type II certifications; clear ownership disclosure; documented operational firewall between hub services and any affiliated specialty pharmacy; defined audit cadence.

Red flag: Specialty pharmacy affiliation with vague firewall documentation, or refusal to provide a BAA pre-contract.

9. Reporting, Analytics, and Data Portability

The reporting baseline is weekly cadence covering enrollment funnel metrics (referrals, enrolled, BV complete, PA submitted, PA approved, first fill), TAT distributions, copay assistance reach, and adherence at 90 and 180 days. Above the baseline, the differentiator is whether you can run your own analysis on raw data or whether you depend on the vendor's dashboards.

Ask: What reports come standard, at what cadence? Can you export raw data into our data warehouse via API or SFTP? What's the data dictionary look like? Who owns the data at contract end? What's the data migration process if we leave?

What good looks like: Defined standard reports, raw data export available, BI tool integration (Tableau, Power BI), clear data ownership and exit migration terms in the contract from day one.

Red flag: Vendor-only dashboards, no raw data export, ambiguous data ownership at contract end. This is one of the most common reasons hub program transitions fail or become costly.

10. Commercial Terms and Exit

Pricing models in hub services typically come in three forms: transactional (per-action pricing for BV, PA, enrollment), dedicated FTE (fixed monthly cost for a named team), or hybrid (FTE plus transactional volume). Each carries different scaling economics.

Ask: What's the implementation cost and timeline? What are the ongoing fees and which actions are billable? What is the change-order policy and is there a cap? What's the term length, and what are the exit and data migration terms?

What good looks like: Transparent unit economics with named billable actions, capped change-order pricing, defined renewal triggers, and a written exit playbook (data migration timeline, transition support, IP terms).

Red flag: Bundled FTE pricing with no transactional benchmark, ambiguous change-order policy, or exit terms that require a separate negotiation at contract end.

The Comparison Table: How the Major Hubs Stack Up

The four largest independent hub services platforms are ConnectiveRx, AssistRx, CareMetx, and EVERSANA. Each has a different profile across the 10 domains. The fifth column is the option that didn't exist five years ago: running your patient services operation in-house with an AI workforce layered onto your CRM.

This comparison is based on public information and aggregates broad market positioning. Buyers should verify any specific claim during their own RFP process.

Use the table as a starting point, not a verdict. Every cell needs verification against a current proposal.

When Should You Build vs. Buy Hub Capabilities?

For most of the last decade, "build versus buy" wasn't a real decision for hub services. Building meant hiring a 40-person team, integrating systems for 18 months, and absorbing the operational risk of a brand-new program. Almost no one chose that path outside the largest pharma manufacturers.

AI workforce platforms changed the math. Operational workflows that previously required dozens of FTEs — outbound payer calls, portal-based PA submission, enrollment intake from fax, foundation searches — can now be deployed as AI workers in weeks. That doesn't mean every manufacturer should run their hub in-house. It means the decision is now worth evaluating on the merits.

The criteria that favor building:

• Patient volume above roughly 5,000 enrollments per year, where the per-action transactional pricing of a full-service hub becomes expensive at scale.

• Internal infrastructure already in place: a CRM, an operations leader, and at least one pharmacist or clinical resource.

• Multiple specialty products in the portfolio, where a shared internal hub amortizes across launches.

• Strong data integration requirements — the manufacturer wants raw operational data flowing into its data warehouse without intermediation.

The criteria that favor outsourcing:

• First specialty launch, no internal infrastructure, runway to launch under 12 months.

• Highly complex therapy (gene therapy, REMS-restricted) requiring deep clinical support and dedicated case management.

• Small volume where building doesn't amortize.

• Capital constraint that makes operating expense preferable to building capability.

The middle path most manufacturers underweight is hybrid. Outsource intake and BV to a hub-lite vendor, then run PA, financial assistance enrollment, and onboarding internally with an AI workforce. Or outsource everything that touches the patient and run the payer-facing workflows internally where the cost-to-serve is highest.

Neon Health works with manufacturers across all three patterns. The deployment looks different — sometimes the AI workforce sits inside an outsourced hub's operations, sometimes inside the manufacturer's own patient services team. The consultative starting point is the same: define the operational model, define the success metrics, then design the AI capabilities to fit.

How to Run the Hub Vendor RFP

A well-run hub RFP takes 10-14 weeks from kickoff to signed contract. Compressing below 10 weeks usually means the manufacturer skipped a step that surfaces later as a change order. Stretching past 14 weeks usually means the operational model wasn't defined upfront and vendors are quoting different scopes.

Weeks 1-2: Define the model and the must-haves. Decide on full-service, hub-lite, hybrid, or in-house-with-AI. Write your three non-negotiables (for example: real-time eBV, ePA-first PA workflow, raw data export to your warehouse). Write your success metrics — TAT median and P95, PA approval rate, copay enrollment rate, abandonment rate, cost per enrolled patient.

Weeks 3-4: Issue the RFP. Structure the RFP around the 10 domains, not around vendor capabilities. Use scenario-based questions: "Walk us through a same-day BV on a $20K specialty therapy with a Medicare Advantage patient who has step therapy." "Show us the PA workflow when payer X requires fax submission and a clinical note." Generic capability questions produce generic answers.

Weeks 5-7: Demos and data exchange tests. Live workflow demos on sample patients, not slide decks. Run a data exchange test — send the vendor a test enrollment file in your preferred format and watch them ingest it. The IT-level friction surfaces here, not in the procurement meeting.

Weeks 8-9: Reference checks. Talk to at least two named programs, ideally one current and one former customer. The former customer is the more informative call.

Weeks 10-12: Pilot. Before committing to multi-year terms, run a pilot program with measurable KPIs against the success metrics defined in weeks 1-2. The CAQH Index identified roughly $20 billion in addressable automation savings across healthcare administrative workflows, with 70 minutes per patient visit saved when those workflows are fully automated [https://www.caqh.org/blog/new-caqh-index-reveals-20b-savings-opportunity-to-cut-waste-reduce-costs-and-improve-patient-access]. Use those order-of-magnitude figures to set your ROI baseline, then measure your pilot against it.

Weeks 13-14: Contract. Lock the data portability and exit terms before signature. This is the leverage point that disappears once the program is live.

Frequently Asked Questions

How long should a hub vendor RFP take?

A well-structured hub services RFP takes 10-14 weeks from kickoff to signed contract, including pilot. Compressing below 10 weeks typically means skipping the model definition or the pilot, both of which surface later as change orders or program issues. Manufacturers with existing vendor relationships often run a faster process for re-bids, but new vendor selections should run the full cycle.

What is the typical contract length for a hub services agreement?

Hub services contracts typically run three to five years, with annual or two-year renewal triggers. The longer term reflects the integration cost on both sides — the manufacturer's data systems and the vendor's operational ramp. Be specific about exit and data migration terms in the contract from day one. The renewal trigger is your leverage point to renegotiate pricing or scope.

Can a small biotech use the same hub vendors as large pharma?

Yes. Most full-service and hub-lite vendors serve manufacturers across the size spectrum, and several specialize in small or mid-size biotech (where the hub is more likely to function as a comprehensive launch partner). Small biotechs should weight commercial flexibility, launch experience in their therapeutic area, and the vendor's willingness to be consultative — the largest hubs are not always the best fit for a 500-patient orphan drug launch.

How is AI changing hub vendor evaluation in 2026?

AI moved from pilot programs to operational technology across hub services in the past 18 months [https://market.us/report/pharma-hub-and-patient-access-support-service-market/]. The evaluation question shifted from "do you use AI" (every vendor does now) to "where in the workflow is AI executing end-to-end, what's the accuracy benchmark, and what's the escalation rate?" Manufacturers should ask for workflow-level specificity, not feature lists. AI workforce platforms also created a real in-house option for capabilities that previously required outsourcing.

What's the difference between a hub vendor and a specialty pharmacy?

A hub vendor coordinates the patient access workflow — intake, BV, PA, financial assistance, onboarding, adherence support — and typically routes prescriptions to one or more specialty pharmacies for dispensing. A specialty pharmacy dispenses the medication and may also handle clinical management, but it sits downstream of the hub. Some hub vendors are owned by or affiliated with specialty pharmacies, which creates compliance considerations around firewalls and steering [https://www.pharmaceuticalcommerce.com/view/patient-support-hub-services-continue-to-evolve].

Should we outsource our hub or build it in-house?

The decision depends on patient volume, internal infrastructure, launch complexity, and time-to-launch. High-volume programs with existing CRM and operations infrastructure increasingly favor in-house operation with an AI workforce. First launches without internal infrastructure typically favor outsourcing. Most manufacturers underweight the hybrid option — outsource intake while running PA and onboarding internally with AI workers — which often delivers the best cost-per-enrolled-patient.

Key Takeaways

• Define the operational model — full-service, hub-lite, hybrid, or in-house — before issuing the RFP. Vendors quoting different scopes produce responses that cannot be compared.

• Score vendors on 10 specific domains, not a generic capability matrix. Weight the domains by what your launch actually needs.

• Demand TAT distributions, not averages. A vendor that won't share P95s has a workflow problem hidden in the variance.

• Treat ePA workflow depth as the leading indicator of time-to-therapy. Manual PA volume disguised as white-glove service produces long tails.

• Validate specialty pharmacy firewall documentation when the hub is affiliated. Independent hubs avoid the question; affiliated hubs must answer it.

• Lock data portability and exit terms in the contract from day one. This is the leverage point that disappears once the program is live.

• AI workforce platforms are now a real alternative for specific capabilities. The build-vs-buy decision is no longer binary, and hybrid models often deliver the best economics.

Closing

The hub vendor you pick shapes time-to-therapy, abandonment risk, and program economics for the duration of the brand. Run the RFP against a real framework, not a checklist that came from the vendor.

The newer reality in 2026 is that "pick a hub vendor" is the wrong frame for many manufacturers. The right question is "what operational model fits this launch, and what mix of partners and internal capabilities delivers it?" An AI workforce can operate inside a full-service hub's operations, inside a hub-lite partner, or inside the manufacturer's own patient services team. The model is the decision; the vendor list follows.

If you're working through hub strategy for a launch or re-bid and want a second opinion from a team that sits on the operations side, Neon Health works consultatively with patient services leaders to map workflows, identify automation candidates, and design the right operating model. Schedule a consultation to walk through your hub strategy.