The productivity crisis in drug discovery has felt insurmountable. But emerging technology can now help organizations rebuild futureproof pharma R&D models.

In the first part of this Pharma Convergence series, The Rise of TechBio and R&D’s Evolving Operating Models, we detailed the data-first mindset now transforming drug discovery. Building on that topic in the second part of this series, Rewiring early drug development with AI and digitization, we examined the potential renaissance in drug discovery being driven by emerging technology (e.g., AI in drug development). But discovery is only the first step in the marathon that is development and commercialization, a process that is fraught with challenges. Despite major technological advances, the cost of bringing a new drug to market has more than doubled since 2013, when it was roughly $800M, to around $2.6B. This paradox, where innovation fails to yield efficiency, has been captured by Eroom’s Law (Moore’s Law, spelled backwards), which observes that drug discovery productivity halves roughly every nine years.

It is important to understand the root causes of this productivity crisis, the emerging technologies that are beginning to reverse the trend (i.e., AI-driven target identification, predictive analytics, and digital trial platforms), and the strategic imperatives (i.e., cost efficiency, accelerated timelines, and sustainability) for pharma leaders to build a more predictive, efficient, and sustainable pharma R&D model. Through the Capgemini lens, we can explore how organizations are navigating this inflection point across the end-to-end discovery and development lifecycle, and how pharma leaders can build drug R&D models that are more predictive, scalable, and future-ready.

The innovation paradox: Why Eroom’s Law continues to shape pharmaceutical R&D  

Despite exponential increases in pharmaceutical research and development spending, the output of new molecular entities (NMEs) has remained flat, and the cost per approved drug has escalated dramatically. Eroom’s Law shows that drug approvals per billion dollars spent have halved every nine years since the 1950s – a trend that persisted through 2024. 

Thought Leadership_productivity gains in R&D - slide 1
Figure 1: More investment but less output

Four catalysts of Eroom’s Law: Dark clouds hanging over drug R&D 

Scannell et al.’s seminal 2012 paper on the decline in pharmaceutical R&D efficiency identified four enduring drivers of Eroom’s Law, all of which continue to put strain on the pharma R&D ecosystem.

Drivers of Eroom’s law:  

  1. The ‘better than the Beatles’ problem: New drugs must outperform highly effective, low-cost generics, raising the bar for clinical evidence and pushing research into difficult-to-treat diseases. 
  1. The ‘cautious regulator’ issue: Heightened regulatory scrutiny, driven by past safety crises and new pricing policies (e.g., the U.S. Inflation Reduction Act (IRA)), has made drug approval more complex, costly, and time-consuming. 
  1. The ‘throw money at it’ tendency: A legacy of equating spending with progress has led to inflated R&D costs and organizational structures, though recent years show a shift toward more disciplined capital allocation. 
  1. The ‘basic research–brute force’ bias: Overconfidence in early-stage, target-based drug discovery led to many lab successes that failed in clinical trials, with predicting real-world efficacy remaining difficult. Pharma R&D productivity also varies widely by therapeutic area, with costs much higher in complex fields like oncology. 
Figure 2: Data highlighting lack of uniformity in R&D productivity (Sertkaya, A. et al 2024) 
Figure 2: Data highlighting lack of uniformity in R&D productivity (Sertkaya, A. et al 2024)

Beyond the hype: Critiquing tech’s role in pharma R&D efficiency 

Technological innovation has long promised to transform pharmaceutical R&D productivity, yet its impact has unfolded in uneven waves. To understand where value has truly been created – and where it has fallen short – it is useful to examine how successive eras of innovation have shaped the way research is conducted. 

The first wave: Industrialization and its limits (c. 2000s): Despite rapid technological advances, pharma R&D productivity has declined. The 2000s saw industrial-scale tools like High-Throughput Screening (HTS), which enabled rapid testing of millions of compounds but did not improve clinical success, as most “hits” lacked relevance to human disease. 

The genomics revolution: Targeting with precision (c. 2010s): The genomics revolution in the 2010s marked a turning point. Today, drugs targeting genes with strong human genetic evidence are over twice as likely to succeed from phase one to approval. For Mendelian disease targets, success rates are six to seven times higher. Biomarker-driven trials, especially in oncology, have increased approval odds up to 8–12-fold and reduced costs and timelines. 

Thought Leadership_productivity gains in R&D -infographic 3
Figure 3: Four factors revolutionizing pharma R&D 

Building the modern stack: AI in drug development and more 

Recent innovation in drug R&D is driven by an “integrated productivity stack.” The following three core technologies are responsible for emerging efficiency gains.  

Transformative tech in pharmaceutical R&D:-

AI and ML:  Now essential for analyzing vast “omics” data, AI in drug development accelerates target discovery, molecule design, and clinical trial optimization. The AI drug discovery market is projected to grow from $3 billion in 2022 to over $16 billion by 2034. AI-driven platforms can halve the time from target discovery to phase two. The ultimate goal of AI is to accurately simulate human pharmacokinetics and pharmacodynamics (in silico) and reduce risky trials. This remains a future goal, but one that is actively being pursued. 

CRISPR:  CRISPR/Cas9 gene editing technology enables rapid, precise gene editing, making preclinical models more relevant and improving target validation. This helps eliminate weak candidates early, saving time and resources. 

Real-World Evidence (RWE):  RWE uses real-world data (EHRs, claims, registries) to inform trial design, speed up patient recruitment, and create external control arms. RWE improves trial design, speeds up patient recruitment by identifying hotspots, and enables external control arms for rare diseases. In 2024, five FDA approvals relied on RWE. The use of RWE unlocks many efficiency gains, including cutting launch research costs by $3 million and reducing time-to-market by 18 months. 

Together, these technologies, built on genomics, create a synergistic stack that enhances prediction, efficiency, and R&D productivity. 

The inflection era for drug R&D: Mapping the impact of innovation on efficiency 

Over two decades, rising drug development costs and waves of innovation have shaped pharma R&D. By mapping these trends against Eroom’s Law, we see a potential inflection point where knowledge-driven technologies are beginning to improve productivity. 

In 2019, an incredible $83 billion dollars was spent on pharmaceutical R&D, which is roughly a ten-fold increase from the 1980s. Today, spending has somewhat stabilized. This shift aligns with the adoption of genomics, AI, CRISPR, and RWE — suggesting early signs of positive impact. 

  • Negative/neutral impact zone (c. 2004-2010): Industrialization (HTS, combinatorial chemistry) increased early-stage activity but failed to improve target validation, driving costs up. 
  • Positive impact zone 1 (c. 2010-Present): Genomic validation doubled the probability of success for genetically supported targets, reducing costly failures. 
  • Positive impact zone 2 (c. 2018-Present): The integrated productivity stack (AI, CRISPR, RWE) began to flatten the cost curve by improving preclinical models, trial efficiency, and candidate quality.  

The below table highlights the significant delayed effect of technology when compared to new discoveries. Though the human genome was sequenced in the early 2000s, its productivity benefits emerged a decade later. Similarly, AI and CRISPR are only now beginning to reduce R&D costs, highlighting the need for long-term investment to realize returns.

Thought Leadership_productivity gains in R&D - slide 4
Figure 4: The delayed effect of technology compared to new discoveries 

The leadership blueprint for pharma R&D: Driving transformation from the top 

Two decades of R&D data reveal a clear path forward: declining productivity will only be reversed when leadership shifts from scaling operations to enhancing knowledge and predictive capabilities.  

Figure 6: Four imperatives to transform pharma R&D 
Figure 5: Four imperatives to transform pharma R&D 

With its deep domain expertise and technology leadership, Capgemini Invent’s Life Sciences team is uniquely positioned to support pharmaceutical enterprises in navigating these transformative shifts. 

References: 

  1.  Jogalekar, A. (2012) ‘The unstoppable Moore hits the immovable Eroom’
  2. Heilbron, K. et al. (2025) Advancing drug discovery using the power of the human genome  
  3. Discovery Life Sciences (2025) Biomarker Strategy Can Increase Your Clinical Trials Success Rate 
  4. Premier Inc (2025) How One Life Science Company Leveraged Real-World Data to Save Money, Time and Support a Regulatory Filing  
  5. Han, H. et al (2025) CRISPR/Cas9 technology in tumor research and drug development application progress and future prospects  
  6. SCW.AI (2025) AI in Pharma: Use Cases, Success Stories, and Challenges in 2025