Clinical trials are run very differently than they were a decade ago. The technology layer around them grew by accumulation rather than by design, and the cost of that is now distributed across everyone in the ecosystem.
Two figures describe how much the execution model has shifted inside a single decade.
Between 2015 and 2025, procedures in phase 3 pivotal trials rose by more than 60 percent, from 187 to 301. Over the same period, investigative sites per phase 3 pivotal trial grew by a similar proportion, from 65 to 106.
More is being asked of each protocol, and the work is distributed across a wider network to deliver it. Neither trend is a mistake. Both follow from real scientific and regulatory developments: more precise patient populations, richer endpoints, broader geographic representation, and higher expectations for the evidence a pivotal study must produce.
But a delivery model spread across nearly twice as many participating organisations behaves differently from one that is not. Coordination cost rises faster than headcount, and the systems layer that holds it together was largely built for the earlier shape.
Each organisation running trials selected systems that suited its own operations. Each choice was defensible on its own terms, made by competent people solving a real problem, and delivered value in the function that chose it.
What none of those decisions could account for is the aggregate. A research site does not experience one organisation's technology stack. A coordinator working across several studies experiences the union of all of them, each with separate credentials, separate training, separate conventions and separate support routes, refreshed with each new protocol.
No one designed that experience and no one owns it. It is an emergent property of a growing market in which every participant optimised locally, which is the normal way ecosystems accumulate friction as they mature.
The effect is now well documented, and the specifics are more surprising than the general finding.
Tufts CSDD research reports that 70 percent of global site staff say trials have become more challenging to manage due to increasing complexity. Sites identify setup and training on study-provided systems as the most burdensome start-up activity they face, ranking above contracting and budgeting. Two-thirds report that technology burdens are greater now than five years ago.
That ordering is worth pausing on. Technology setup outranks contract negotiation as a start-up burden, in a period when technology was expected to reduce administrative load rather than add to it.
The Society for Clinical Research Sites has found that sites strongly prefer integrated systems and are calling for platforms that work across multiple studies and organisations. That preference is rational, and it is also structurally difficult for any single organisation to satisfy, since none of them controls what the others deploy.
It matters that this is a systems problem rather than an attitude problem, because the two invite different responses.
If fragmentation were caused by insufficient attention to the site experience, the answer would be better engagement programmes, more advisory boards and improved communication. Those things have value, and they have been deployed widely, and the burden figures have continued to rise.
The reason is that the friction is created by the number of boundaries in the system, not by the quality of relationships across them. Every additional platform adds a credential set, a training obligation, a support path, a data reconciliation point and a version of the truth that has to be kept consistent with the others. Goodwill does not reduce any of those. Architecture does.
This also explains why the ecosystem-wide version of the problem is hard. Coordinating platform choices across independent organisations is a collective action problem, and the industry has no mechanism for it. What each organisation can address is its own contribution, and that portion turns out to be substantial.
The case for consolidation is usually made on internal grounds: integration cost, data consistency, validation overhead. Those are real, and they understate the effect, because the benefit lands on several parties at once.
For the organisation running the study, one platform means one validation boundary rather than several, one integration surface, and a single data model in which enrollment, monitoring, documentation and payments reference the same records. Questions that previously required joining exports across systems become reports.
For CROs and partners working across a portfolio, it means fewer environments to staff, train for and maintain, and less effort spent reconciling data whose only inconsistency is that it came from different places.
For research sites, it means fewer credentials, one training path and one support route per study rather than one per system, and less duplicate data entry. This is the part most directly connected to the burden figures above.
Cloudbyz brings CTMS, eTMF, EDC, RTSM, site payments and the site-facing portal onto one platform built natively on Salesforce. The practical consequence is not a new capability so much as the removal of the seams between existing ones, which is where most of the accumulated cost actually sits.
Most organisations assess the site experience through relationship surveys. These capture sentiment well and diagnose causes poorly, and the large annual benchmarking exercises, valuable as they are, arrive too late and too aggregated to drive a specific change.
A more useful operational metric is a burden count, maintained per study. Number of site-facing systems. Number of distinct credentials. Number of required training modules. Estimated hours of site time consumed before first patient in. Number of duplicate data entry points across the study.
It takes an afternoon to compile, it is comparable across studies, and it converts a general concern into a number a study team can be asked to reduce. A satisfaction score reports that something is wrong. A burden count identifies what to remove.