
Selling white coats software
There’s broadly two kinds of bio Saas: owning the data layer / orchestration of the most vital functions and selling subscriptions for discovery tooling
Veeva, Certara, Schrödinger, Simulations Plus, Benchling, Dotmatics, Latch Bio, Unlearn AI, Vial, Convoke, Sleuth, Medidata, Noetik, Chai, Boltz
Compound portfolio companies Briefly and Spaero
Systems of Record: Data Layers & Orchestration of the Most Vital Functions
These companies modeled more closely to traditional tech SaaS are currently somewhat out of favor given that Benchling is the only recent startup that has been a qualified successes. However, people easily forget about companies like Veeva that ramped to $2.8B in revenue at 75% gross margin as detailed below.
Just like in traditional tech SaaS, the prime “beach front” SoR properties of CRM, clinical trial & EMR management command orders of magnitude more value than all other areas. While the aforementioned companies currently own those properties, we view the likes of Veeva, Certara, etc. as being very much surmountable (unlike Epic Systems in healthcare or Salesforce, etc. in traditional SaaS).
We believe this not only because their positions aren’t all that strong but also because biotech is more gated than traditional SaaS by intelligence (e.g. digesting academic literature) and interpersonal interaction (e.g. 40 minute calls with thousands of patient for trial recruitment). Indeed, the largest biopharma SaaS companies in existence are still often 60% consulting-like services businesses vs 5-15% for the tech SaaS giants. AI is of course the fixer.
AI may well be enough to dislodge the incumbents holding the most desirable properties because the way the technology is built shifts from the UI / front-end being as easy as possible for humans to manually enter and edit data. Going forward, the winner might be the builder of the most flexible and efficient database for use by AI agents.
We’re also particularly excited about startups automating clinical trial recruitment with the ultimate goal of being an end-to-end clinical trial operator. The company’s core competency beyond cheap operation should be optimizing for each individual trial the most efficient and effective possible route with modern tools: integrate the short-cuts to getting human evidence as detailed in Shelby’s fantastic post and biomarkers.
Drug Discovery Tooling
Companies building here should be keenly aware of the difficulties of selling pre-clinical tools to pharma. Pharma is very reluctant to try novel technologies and change their workflows around them. Worse, they will pay minuscule amounts for it.
As such, discovery-focused companies typically pursue the platform route commercialize via services and JVs where they orchestrate their own tech to help pharma hit targets that have proven impossible to drug with their existing internal methods, which is of course a high bar.
It normally works up from services payments to JVs after customer feels the value and the startup can negotiate for larger downstream economics.
Building a business purely around subscriptions is so much harder. Schrödinger is the most successful version. They launched in 1990 under what sounds like a paradigm-shifting premise of applying computation to drug discovery, chiefly via MD simulation. 25 years later, they make $200M in revenue and have turned a profit in just one year. They’ve now pivoted to also doing internal drug development (in addition to spinning off Nimbus).
This industry dynamic of pharma paying pennies for pre-clinic discovery tools may change if and only if pharma starts to view novel approaches as existential threats. We’re currently seeing this play out with our portfolio company Wayve. Just within the last year, the legacy car OEMs have internalized that AVs are here and are existential to their very survival. Negotiations have shifted from contracts for tiny margins won via hand-to-hand combat over 2-year sales cycles to truly massive deals with 3 month sales cycles.
SaaS startups building discovery tooling especially expensive tooling like foundation models are implicitly betting the company on their tech striking sufficient fear into pharma such that it trigger its survival instinct such that they can get 8+ figure up-fronts for pre-clinical tooling or JVs.
Early in 2026, we may be seeing the start of this for AI models with Noetik, Chai, and Boltz all securing large software license deals with pharma. Several are $10M+ per year without exclusivity.
One final thing to consider is that hiring a sales/BD team so strong that they can keep going a perpetual pipeline of software deals worth $10s-100s of millions isn’t that much easier (if easier at all) than hiring a great drug development team and pursuing partnerships and/or an internal pipeline. The difference is that historically the latter captures orders of magnitude more upside if it works.
And, hiring quality BD people into SaaS biopharma startups is hard and more importantly they’re too expensive for early stage startups and don’t tend to be the right cultural fit (e.g. they need EAs to do everything, want a company car, etc.).
This tends to mean that these startups rely on founder-led sales. Deciding if this is a good fit requires a deeply personal assessment of themselves, their desire to devote themselves to B2B sales growth, and depth of personal network in pharma BD teams.
For more on selling software to biopharma, read this excellent post.
https://astera.org/software-sales-handbook/
https://www.aibiodesign.com/p/selling-ai-products-services-to-big
https://shawndimantha.substack.com/p/the-techbio-idea-maze-to-be-or-not
https://centuryofbio.com/p/vial
https://scuttleblurb.substack.com/p/veeva-ai
https://www.benchling.com/blog/biotech-guide-to-data-driven-rd
https://www.michaeldempsey.me/blog/2025/10/03/sequencing-vs-equal-odds-applied-research/
Contract sizes for scaled companies:
| Company | “Typical” individual contracts (what’s observable) | NRR / retention (software) | Typical buyers | Customer count |
|---|---|---|---|---|
| Veeva (VEEV) | Deals range widely (from <$250k single modules to multi-million platform suites). Independent purchasing data show median Veeva contract ≈ $211,872/yr (limited sample). Mgmt flagged a top-20 pharma multi-app clinical platform win as one of its largest subscription orders ever. | Veeva doesn’t publish a current NRR, but historically disclosed subscription services revenue retention of 119–124% (FY20–22). | Global biopharma across sizes (top-20 to emerging biotech); also medtech and sites for SiteVault. | 1,477 customers as of FY25 year-end. |
| Schrödinger (SDGR) | Very transparent ACV tiers: Total ACV $190.8M (2024); 31 customers ≥$1M ACV, 8 customers ≥$5M ACV. Across 1,752 active software customers, implied avg ACV ≈ $109k (blend of academics, biotech & pharma). | Reports 100% retention for customers ≥$500k ACV (2024); 95% for ≥$100k ACV. | Enterprise biopharma R&D; mid-size biotechs; 1,250+ academic orgs historically; some materials science. | 1,752 active software customers (ACV ≥$1k) at 12/31/2024. |
| Certara (CERT) | Mix of seat licenses (Phoenix, Simcyp), enterprise validation (Pinnacle 21), etc. Marketplace data: avg Certara software buy ≈ $16k/yr (max ~$41k) for smaller purchases; large enterprise footprints scale higher. | NRR (software) 114% (Q1’24); 102% (Q1’25), rebounded to 107.6% (Q2’25). | Large pharma, biotechs, CROs; regulators (FDA, PMDA) use Pinnacle 21; broad R&D/clinical functions. | 2,400+ organizations (biopharma, academia, regulators). |
| Simulations Plus (SLP) | Software avg revenue per customer ≈ $129k/yr (FY2024). | Renewal (software): ~90% by fees (TTM), ~84% by accounts (varies by quarter). | Pharma & biotech teams doing PBPK/QSP/PK-PD; some regulatory & academic users. | Not regularly disclosed; serves hundreds of biopharma customers globally. |
| Medidata Solutions (MDSO; acquired by Dassault Systèmes—completed Oct 29, 2019) | Multi-study & single-study subscription arrangements; subscription term generally 1–5 years. SEC Disclosed 12-month subscription backlog ~$522M and total multi-year subscription backlog ~$1.17B at 12/31/2018. SEC+1 FY2018 subscription revenue $535.7M (useful for “implied average” spend calculations). SEC | Revenue retention rate >99% (2016–2018). SEC Q2’19: retention reported as nearly 100%, and Medidata defines the metric as the % of prior-year revenue attributable to customers retained in the current year. SEC+1 | Pharma/biotech/medical device & diagnostics companies + institutions (academic/government/non-profit), CROs, and other clinical-trial sponsors/operators. SEC+1 | Over 1,000 customers as of 12/31/2018. SEC 1,330 total customers as of Q2 2019 (press release furnished via SEC). SEC+1 |
Veeva’s GTM & Distribution
Schrödinger GTM & Distribution