A Conversation with Elizabeth Iorns - Advice for Biotech Founders

Elizabeth Irins: Background & Origins

• Holds a PhD in cancer biology from the University of London; served as an assistant professor at the University of Miami School of Medicine until founding Science Exchange.
• Founded Science Exchange (originally named "The Bench") in 2011 as a Y Combinator Summer batch company, identifying as the first biotech-like startup in that cohort.
• Currently serves as Chairman of Reformer Therapeutics (also a YC company) and is a part-time partner at Y Combinator advising hundreds of biotech firms.
• The startup idea emerged from personal inefficiency in academic research, specifically the difficulty of sourcing specialized experiments, evaluating provider quality, and managing intellectual property rights.
• Modeled the business on B2B freelance marketplaces like Upwork (then Odesk/Elance) to solve the fragmentation in scientific collaboration infrastructure.

The Science Exchange Marketplace Model

• Addresses a market where price variance for identical experiments can reach 10x due to information asymmetry and lack of standardized data.
• Operates as a curated B2B marketplace rather than a consumer platform, requiring rigorous Quality Assurance (QA) to qualify and contract every external provider.
• First product-market fit (PMF) was achieved when Amgen launched the platform for its entire discovery research sector, shifting from manual vendor management to the platform.
• Primary competitor is identified as the "status quo" of internal, manual vendor management processes (e.g., complex SharePoint setups) rather than other software firms.
• First revenue came from a "cold" user running a microarray experiment while the founders were attending a wedding in Italy; initial users were personally recruited friends.
• Early growth strategy involved a "concierge" MVP approach to validate demand before building automated infrastructure.
• Successfully onboarded large Contract Research Organizations (CROs) by demonstrating the platform's ability to manage project lifecycles for all external partners, not just one-off orders.

The Reproducibility Initiative

• Launched the Reproducibility Initiative to validate published scientific results, driven by the hypothesis that a network of vetted labs is necessary to replicate experiments efficiently.
• Projects include antibody validation, epidemiology reanalysis (with the Gates Foundation), and public replications of cancer biology and prostate cancer results.
• High-profile public replications are controversial because they challenge the academic norm of not publishing negative or failed replication results.
• Data from the initiative suggests the majority of published results are not reproducible, largely due to poor assay validation and lack of Standard Operating Procedures (SOPs) in academia compared to pharmaceutical industry tech transfers.
• The initiative is credited as a "game-changer" for Science Exchange, significantly boosting brand trust and unlocking major pharmaceutical partnerships despite being a non-revenue-generating mission.

Current Challenges & Scaling

• The primary challenge for the CEO shifted from acquiring initial users to scaling operations while maintaining discipline with a team of 85 people handling large enterprise integrations.
• Science Exchange manages over $100 billion in annual outsourced research spend within the pharmaceutical industry.
• The platform achieves a Net Promoter Score (NPS) of 78 for clients and 67 for suppliers, compared to an industry average of zero, attributed to standardized deliverables and transparent performance data.
• Scaling strategy focuses on becoming the central system of record for external research partners rather than a transactional tool.

Biotech Industry Trends & Advice

• The biotech sector is experiencing an unprecedented influx of capital and a "Moore's Law" moment regarding therapeutic modalities (e.g., recent approvals of gene therapies, cell-based therapies, and RNAi).
• A new trend involves industry veterans with drug commercialization experience founding early-stage biotechs, a shift from the previous dominance of purely academic founders.
• Unlike software, biotech cannot pivot around scientific failure; success depends on hitting specific milestones that de-risk the science.
• More biotechs are now commercializing their own products (building sales forces) rather than exiting via acquisition, driven by easier FDA approval pathways for rare diseases and smaller trial sizes.
• Common founder mistakes include avoiding the "killer experiment" (the minimal test that could kill the hypothesis) due to fear of failure.
• Academic founders often face IP hurdles when spinning out research; advice suggests completing grant-funded R&D before joining Y Combinator to minimize ownership disputes.
• Non-scientist founders can succeed in biotech by demonstrating deep domain research (surrogate credentials) and personal motivation, though a PhD is not strictly required if the team includes strong scientific leadership.
• Key advice for non-scientists entering biotech includes self-education to match expert knowledge levels or focusing on non-lab roles like bioinformatics.
• Future biotech products will increasingly rely on "user-pay" models where patients are willing to pay for treatments for debilitating conditions (e.g., migraines) rather than preventive or mild-symptom therapies.

Founder & Team Dynamics

• Elizabeth Irins admits she initially hesitated to leave academia due to fears regarding university ownership of ideas, but the Tech Transfer office was uninterested, facilitating her departure.
• Her boss, the Dean of Medicine, provided critical support by letting her take a sabbatical, contrasting with the common experience of mentors actively blocking academic entrepreneurship.
• For founding teams, Irins prioritizes finding co-founders who "hustle" and share a deep care for the problem over those with specific business titles; she hired a CFO only when reaching a growth stage.
• Incentive structures in the marketplace are driven by the transparency of provider performance data, which penalizes poor performance across the entire user base.