How CROs reshaped biotech and why it matters today

Biotech today runs on an interconnected web of partners: CROs, CDMOs, academic labs, and startup biotechs all working in tandem. But it wasn’t always this way.

In the 1990s, large pharmaceutical companies did almost everything themselves. Drug discovery, clinical trials, regulatory submissions, manufacturing – all in-house. That model, known as the Fully Integrated Pharmaceutical Company (FIPCO), dominated for decades. 

Over the past 30 years, that paradigm has totally unraveled. In its place emerged a modular, outsourced system, powered by specialized contract organizations that now play core roles in drug development. This evolution has powered the rise of lean biotechs, accelerated development timelines, and created a global pipeline of innovation. Here’s how we got here, and why it matters today.

Phase 1: FIPCOs and the rise of CROs (1980s–1990s)

In the early 1990s, less than 5% of pharmaceutical R&D spend went to external providers. Big Pharma ran everything internally. But the seeds of outsourcing were already being planted. In the late 1970s and early 1980s, companies like Quintiles (founded in 1982), Parexel, and PPD began offering services to pharma companies, initially focused on clinical trial biostatistics and monitoring.

At first, CROs were niche. They filled capacity gaps – running a trial here, managing data entry there. But some forces converged to change that:

• The blockbuster drug boom of the late 1980s and 1990s brought enormous revenue potential, but also ballooning R&D budgets and larger, more complex trials.
• The 1996 introduction of the ICH E6 Good Clinical Practice (GCP) guideline was a turning point. For the first time, there was a harmonized global standard for how trials should be designed, run, and monitored. The bar for quality and consistency was raised – and sponsors needed expert partners who could operate under these new expectations.

CROs, with their growing networks and operational expertise, became essential allies. They had trial templates, site relationships, and systems that worked.

By the late 1990s, CROs weren’t just support players, they were infrastructure. Full-service trial management became the norm, and Big Pharma started shedding internal teams and expanding relationships with contract partners.

Phase 2: How outsourcing went from service to infrastructure (2000s)

The early 2000s marked a turning point. CROs grew in size and scope, offering end-to-end development capabilities. At the same time, CMOs and CDMOs began to emerge as critical players –  and manufacturing, long considered core IP territory, slowly became externalized. Why?

• Cost pressure. Maintaining in-house trial ops and manufacturing facilities became unsustainable, especially as pipelines thinned.
• Manufacturing overcapacity. Empty plants became liabilities. Firms like Patheon, Lonza, and Catalent bought these facilities, offering production capacity on demand.
• The patent cliff. As top drugs lost exclusivity, pharma firms downsized and shed non-core assets.
• Venture-backed biotechs surged post-2000, bringing innovation and capital but lacking internal infrastructure. CDMOs filled that gap.

Patheon (later acquired by Thermo Fisher) and Lonza began buying underutilized manufacturing facilities from pharma companies. These contractors could produce drug substance and product more efficiently and flexibly, especially for small biotechs that didn’t have plants of their own.

CROs like Quintiles expanded into preclinical research, pharmacovigilance, and health economics. By 2005, sponsors could outsource nearly the entire development pipeline. Pharma companies could theoretically bring a drug to market without touching a pipette or owning a factory.

Phase 3: The creation of modular biotechs  (2010s)

As CRO/CDMO capabilities matured, a new operating model emerged: the modular biotech.

Instead of building labs or hiring dozens of scientists, startups built tight core teams - program managers, scientific leaders, and BD - and outsourced everything else. CROs synthesized compounds, ran experiments, and managed trials. CDMOs handled formulation and production.

Companies like Nimbus Therapeutics helped popularize the model. At its peak, Nimbus had fewer than 20 full-time employees but managed multiple development programs with support from a network of 80+ CRO scientists. Their internal costs were 10–15% of total spend, with the rest flowing to outsourced work. The benefits were obvious:  

• Capital efficiency: no need to fund wet labs or build teams prematurely.
• Flexibility: ramp spend up or down by project stage.
• Specialization: hire the best experts for a specific task, even if only needed for a month.

In parallel, academia increasingly focused on early discovery and translational research. Universities spun out biotech startups, which then turned to CROs and CDMOs to execute. Academic Research Organizations (AROs) that had pioneered large trial designs handed off that infrastructure to CROs, which scaled it commercially.

By the late 2010s, it was clear: the modular biotech model wasn’t a workaround. It was a new operating system.

Phase 4: Outsourcing as an expected strategy  (2020s)

Fast forward to today: CROs and CDMOs are embedded collaborators, and it’s pretty much expected for any biotech - especially early-stage - to outsource R&D to them.

$82 billion was spent globally on outsourced R&D work in bio in 2022, and that number is expected to pass $135 billion by 2029. The top CROs and CDMOs run global networks capable of taking programs from IND to commercial launch – meaning biotechs can now assemble networks of specialized CROs: one for toxicology, another for assay development, a third for Phase I trials, and so on. 

Even Big Pharma borrows from the playbook. Project teams often manage external portfolios, running multiple CROs/CDMOs in parallel. The drug development pipeline today resembles a neural network more than a factory line.

As a result, the shift has now moved from should we outsource to how do we manage the network of outsourced work – which leads us to…  

Phase 5: Tools to coordinate the network sprawl 

Outsourcing creates leverage, but also complexity. And complexity, left unmanaged, becomes friction.

To make the modular model work at scale, companies need to operate like systems thinkers:

• Run external vendors like an extension of your internal teams
• Centralize decision-making data so you’re not chasing updates across emails
• Build workflows where data, progress, and accountability are shared in real time

That’s where Kaleidoscope comes in.

We’re building the infrastructure layer to make modular biotech run smoothly. We’ve seen how CRO and CDMO coordination breaks down when timelines, data, and decisions live in separate places. Usually, these breakdowns go well beyond a file being shared late or in the wrong email thread – it comes down to rigorously preserving context, version control, audit trails, and visibility. 

We help biotechs centralize their R&D data across the network of internal teams and external partners, so that everyone is operating off the same playbook – whether you’re running your first IND or juggling five parallel programs. Think GitHub for R&D workflows, but built with regulatory-grade security. 

Our reason for building this is clear: history shows the companies that treat CRO/CDMO networks as strategic infrastructure, not transactional vendors, move faster, make better decisions, and scale more effectively.

But while the modular model works, it also requires a totally new operating mindset to manage and coordinate the “new normal”. And that has to start with tools and systems built for this way of work.

Kaleidoscope is a software platform for biotechs to robustly manage their R&D operations. With Kaleidoscope, teams can plan, monitor, and de-risk their programs with confidence, ensuring that they hit key milestones on time and on budget. By connecting projects, critical decisions, and underlying data in one spot, Kaleidoscope enables biotech start-ups to save months each year in their path to market.