Pricing: a platform fee plus a success fee per randomized patient

How does Bond's pricing work?

There are two fees. The platform fee is volume-based and covers EHR integration, implementation and ongoing operation, itemized below; there is no separate integration fee. The success fee is charged per enrolled patient, which Bond defines as a patient who is successfully randomized.^⁠[1]

The split follows the costs. Connecting an EHR, passing a security review, configuring scripts and running the platform are covered by the platform fee. The success fee is the part that moves with results.

What does the platform fee cover?

• EHR connection. FHIR or HL7 interfaces, or a connection through an aggregator, set up by Bond with no separate integration fee. Full EHR integration typically takes 48 hours, depending on the EHR, IT review and interface method; the implementation page shows the plan step by step.^⁠[1]
• Security review and audit logging setup. Bond goes through your security review, signs a business associate agreement (BAA) and sets up audit logging. Details are on the security page.^⁠[1]
• Workflow configuration. The study's inclusion and exclusion criteria and the outreach scripts, configured for each site and study.^⁠[1]
• Ongoing operation. Compute, monitoring and support for as long as the engagement runs.

What is included in every engagement?

• LLM-based EHR screening in Identify, with criterion-to-evidence rationale for each match
• Voice and SMS outreach, pre-screening and scheduling in Engage
• AI consent support in Consent; the site and PI still obtain consent
• Real-time dashboard and audit trail
• EHR vendor integration via FHIR
• Dedicated support^⁠[1]

What does the fee pay to run?

Bond reads structured and unstructured EHR records against the study's inclusion and exclusion criteria and ranks candidates. Bond reports 90%+ matching accuracy for eligibility screening, and each match still shows the evidence behind each criterion, so a coordinator can check the reasoning before acting on it.^⁠[1] Voice and text agents then run outreach with the scripts configured for that site. Patients are told AI assistance is used and can reach a person at any time: the agent transfers the call live to a coordinator or books a human callback, whichever the site prefers.^⁠[1],[2]

PHI is handled under a signed BAA, with encryption at rest and in transit (AES-256 where applicable), role-based access control, SSO support and audit logging.^⁠[1]

What counts as an enrolled patient?

Bond defines enrolled as successfully randomized.^⁠[1] A match, a referral, a completed pre-screening call, a booked visit or a signed consent form does not count. A patient who consents and then fails screening generates no randomization fee.

Randomization is an event every randomized study already records, so the count can be checked against the study's own records. Where a study needs it, milestones for individual visits can be added on top, and those are written into the contract.^⁠[1]

The gap between a referral and a randomized patient is wide. In a vendor-sponsored survey of 100 US site professionals, fielded by myTomorrows in May 2026, 52% said at least one in five referrals they receive is ruled ineligible before formal screening begins.^⁠[7] In one late-life depression trial in Toronto, 4.3% of Facebook self-referrals were randomized (14 of 323), against 30.3% of provider referrals (10 of 33).^⁠[8] That is one small trial, but it shows how far a per-referral count and a per-randomization count can diverge.

The coordinator sees the same funnel the fee is based on. The dashboard reports patients matched, contacted, pre-screened, consented and randomized, along with time to enrollment and screen-failure signals. Reporting is available for sites, CROs and sponsors.^⁠[1]

How does this compare with other ways to pay for recruitment?

Each common way to buy recruitment puts the risk of a slow study in a different place. The table compares structure, not price.

Media recruitment reaches people who are not in the site's records at all. Bond covers that path too: it sets up Meta and Google ad campaigns for your studies, and those leads go through the same pre-screening and scheduling as EHR matches.^⁠[2] The cost of media recruitment varies widely. A 2026 Tufts CSDD study of 32 studies from eight sponsors and CROs found a median centralized outreach budget of $1.33 million, and median outreach cost per patient ranging from $143 in vaccine studies to $11,392 in immunology studies.^⁠[9] The abstract does not say whether a patient means referred, enrolled or randomized, and the figures cover centralized outreach spend only.

The evidence cuts both ways. A 2020 meta-analysis found online recruitment cheaper per enrolled participant (median $72 against $199 offline), but offline recruitment turned screened people into enrollees at a higher rate. The authors suggest one reason for the second finding: sites already hold health records of suitable patients.^⁠[10] For a fuller comparison, see Bond vs media recruitment and recruitment software compared.

What does a site that does not enroll cost?

Tufts CSDD's analysis of more than 150 studies and nearly 16,000 sites found that 11% of sites in a given trial typically fail to enroll a single patient, 37% under-enroll, 39% meet their targets and 13% exceed them.^⁠[11] The data are global and were published in 2013, so treat them as a baseline rather than a current rate.

Every activated site carries fixed costs, whether or not it enrolls. In oncology, a 2023 analysis by the data company Phesi of 173 cancer trials and 11,826 sites found that 19% of sites enrolled just one patient. Phesi estimated that those sites cost about $130,000 per patient, against $14,167 at a better-performing site.^⁠[12]

Slow sites also stretch the timeline. The $55,716 daily Phase III figure from Tufts CSDD covers direct trial costs only, before any lost sales.^⁠[13]

A site that enrolls no one still costs the sponsor money, and costs the site the coordinator hours spent looking. A success fee tied to randomization does not fix that by itself. What it does is keep the variable part of Bond's price at zero for a site that does not enroll.

Where does recruitment cost sit, manually and with Bond?

When a site recruits from its own records, the main cost is coordinator time, and it is spent whether or not anyone is randomized. With Bond, part of that work moves to software covered by the platform fee, and part of the price waits for randomization.

What does Bond not do?

• It does not publish a price list. Quotes are custom and scoped to your sites and studies.^⁠[1]
• It is not purely pay-for-performance. The platform fee is owed whether or not anyone is randomized. Only the success fee, plus any visit milestones agreed in the contract, depends on results.^⁠[1]
• It does not guarantee enrollment. A protocol with few eligible patients, or a site without visit capacity, will enroll slowly with any tool. The success fee means Bond shares that outcome. It does not remove it.
• It does not obtain consent or decide eligibility. The PI and delegated staff do both.

How do you get a quote?

Book a demo and bring one protocol. To scope the platform fee, Bond needs to know which EHR each site uses, how many sites and studies are in scope, whether a study will use Meta and Google ad campaigns, and whether you want to start with a pilot that does not need the EHR connection.^⁠[1],[2]