The Benchling success story starts with a very specific kind of frustration. Not the consumer kind, where an app is slow or a checkout is annoying. The scientific kind, where you spend years training to do work that genuinely matters and then discover that the infrastructure supporting that work is stuck in 1995.
Sajith Wickramasekara grew up in Raleigh, North Carolina, the kind of kid who spent lunch breaks in the high school computer lab because that’s where the interesting problems were. He worked in biology labs during high school, running experiments, collecting data, understanding at an early age how science actually got done day to day. Then he went to MIT to study electrical engineering and computer science and kept working in wet labs on the side.
What he saw there stayed with him. Scientists doing some of the most sophisticated biological work in the world, tracking everything in paper notebooks and Excel spreadsheets. Experiment records that couldn’t be searched. Data that was siloed on individual laptops. Version control that meant printing a new copy and hoping you remembered which one was current. The biology was extraordinary. The software layer underneath it was not.
He left MIT on a leave of absence in 2012 at 22. He has not gone back.
The Problem Nobody Was Building For
The gap Sajith identified was not obvious to the people who controlled capital at the time.
He went to software investors. They looked at biotech and thought the market was too small and too specialized. He went to science investors. They understood R&D pain but had never funded software companies. They invested in drugs, not tools.
Benchling sat in the exact middle of two investor categories that didn’t overlap, which meant early fundraising was nearly impossible. Nobody had a framework for a company that was unmistakably a software business and unmistakably a life sciences business at the same time.
Y Combinator was the exception. Paul Graham understood the breadth of the problem even if the vertical was narrow. Benchling joined the Summer 2012 batch. The $150,000 and the network were less valuable than the validation that the idea was worth pursuing. That credibility helped Sajith recruit engineers from Google, Meta, and Palantir who had biology backgrounds alongside their software credentials, exactly the hybrid profile the company needed.
The co-founder pairing mattered too. Ashu Singhal, who had interned at Google and done research at MIT’s CSAIL, brought the technical depth that matched Sajith’s product vision. Together they built for a customer most software people had never spent time with: a bench scientist doing experimental biology in a wet lab, running assays, designing DNA sequences, managing samples across months-long experiments.
That customer was not hard to find. The first users were people they knew from MIT. After that, Sajith and Ashu would drive to UC Berkeley and Stanford, walk from lab to lab, and sit down with scientists to understand exactly where the software friction was. Not demos. Listening sessions. Where does this break? What would have to be true for you to actually use this every day?
The Free Academic Bet That Everyone Called Crazy
The most important early strategic decision at Benchling was also the one that made the least financial sense in the near term.
They gave the product away to academic labs for free. Not a time-limited trial. Free. Ongoing. As long as you were a researcher at a university or research institution, the full product was yours at no cost.
People told them this was crazy. Academic labs have no budget. Scientists are notoriously resistant to changing workflows they’ve been trained on. There’s no freemium funnel where academics suddenly start paying. You’re building a user base that will never convert.
Sajith and Miles Grimshaw at Benchmark, an early investor, held their nerve on this anyway. The logic was not about academic revenue. It was about something harder to measure and more durable: building a moat around the population of scientists who were learning to do biology and who would carry their tool preferences with them wherever their careers took them.
The analogy to GitHub is useful here. GitHub gave away hosting to open source projects and academic users for years. The engineers who learned Git workflows in college brought those workflows into companies when they graduated. The adoption of the tooling inside the professional population was seeded by free adoption in the academic population that fed into it.
Benchling ran the same play with a longer feedback loop. PhD students and postdocs who used Benchling at Stanford or MIT graduated, joined biotech startups or pharma companies, and brought the product with them. “We want to use Benchling” became a request from scientific talent rather than a sales pitch from a vendor. The friction of enterprise sales got reduced dramatically because the buyer already trusted the product and the seller was often someone inside the organization.
Wickramasekara described it simply: their love for the product as end users gave Benchling the shot on goal. Not the sales deck. Not the enterprise contract. The fact that a scientist already knew how to use it and preferred it over whatever the incumbent was.
This strategy required patience that most venture-backed companies do not have. Benchling was building its go-to-market in the academic ecosystem for years before commercial revenue justified it. The foundation compounded slowly and then very fast.
What the Product Actually Does
Explaining Benchling to someone who has never worked in a biology lab is the same challenge Benchling has had describing itself to investors since 2012. The product is genuinely complex because the science it supports is genuinely complex.
Start with the simplest piece: the electronic lab notebook. A scientist records an experiment, the materials used, the protocol followed, the results observed. On paper this is a literal notebook. In Excel it’s a spreadsheet with no structure, no version control, and no way to connect to the samples or sequences the experiment was about. In Benchling it’s a structured, searchable, linked record that connects to every other relevant piece of data in the system.
Then add molecular biology tools. DNA sequence design and editing. CRISPR guide RNA design. Plasmid mapping. The ability to design a genetic construct in the same system where you’re recording the experiment that will use it. For biotech companies whose entire workflow depends on designing genetic sequences and testing them, having design and documentation in the same platform rather than siloed across separate tools is not a minor convenience. It changes how the whole workflow operates.
Then add sample management. A research project might track thousands of samples across months, stored in different freezers, processed in different workflows, referenced in dozens of experiments. Benchling’s registry is the system of record for all of it, with lineage tracking that lets a scientist trace exactly what happened to any sample at any point.
Then add workflow management. Coordinating tasks across specialized R&D teams, capturing data automatically from lab instruments, standardizing protocols so experiments are reproducible across different scientists.
Then add the compliance layer. Benchling’s Validated Cloud, launched in 2021, supports GxP requirements, the Good Practice regulations that govern pharmaceutical and clinical manufacturing. This is the piece that let Benchling move from early research into development and manufacturing workflows, dramatically expanding the addressable market within each customer.
The GitHub comparison is imperfect but instructive for the structure. Biology research has code: DNA sequences, experimental protocols, molecular designs. It has version control problems. It has collaboration problems. It has the same need for a shared, searchable, authoritative repository that software development had before Git. Benchling is building that infrastructure layer for biology.
Four Straight Years of Doubling
Between 2017 and 2020, Benchling doubled its ARR every single year. Four consecutive years. That is a compound growth rate that looks absurd on a chart and requires both an excellent product and a sales motion that is working.
The sales motion was working because of the academic foundation. Scientists who had used Benchling in graduate school were now at Gilead, Sanofi, Regeneron, Corteva. When Benchling’s sales team knocked on the door of a pharma company, there were often already dozens of scientists inside the building who had been using the product for years and wanted their company to adopt it officially.
The pandemic accelerated everything. COVID-19 was the moment where the world suddenly cared about how fast biology could move. Benchling’s customers working on antibody treatments estimated that the platform shortened their time to deliver those treatments to market by 70%. That number got attention. Organizations that had been slow-walking decisions about modernizing their R&D infrastructure suddenly had urgency.
Benchling doubled ARR again in 2020. That was the fourth straight year. By early 2021 the company was adding customers like Gilead Sciences, Sanofi, and UCB. The Series E in April 2021 brought in $200M led by Sequoia. Sequoia’s framing was explicit: biotech breakthroughs have the potential to transform the world the way computing and the internet did, and Benchling was building the foundational software infrastructure for that transformation.
The valuation reached $6.1 billion following a subsequent financing round led by Franklin Templeton and Altimeter. It was the kind of number that, for a company most people outside of biotech had never heard of, required explaining why the market was larger than it looked.
Why the Market Is Larger Than It Looks
The pharmaceutical industry spends over $200 billion annually on R&D. By most estimates, a substantial fraction of that spending is on administration rather than actual science. Scientists spending time managing spreadsheets, reconciling disconnected records, searching for data that exists somewhere but is not findable. The McKinsey estimate that 45% of major diseases will be addressed by biology over the next decade, with biotech potentially driving $2-4 trillion of economic impact annually, frames the size of what is being built on top of the software layer Benchling is building.
The biotech data volume problem is also accelerating. In 2024, the volume of biotech data was doubling roughly every seven months. The shift from small-molecule drug development, where chemistry tools have been digital for decades, toward biologics, cell therapies, gene therapies, and mRNA therapeutics, creates entirely new categories of complex data that existing tools were not built to handle. A CAR-T cell therapy generates a fundamentally different kind of data than a traditional pharmaceutical compound. An mRNA vaccine manufacturing process has a fundamentally different data structure than a tablet formulation.
Benchling was designed for the new biology, not adapted from tools built for the old one. That architecture advantage matters when customers are increasingly building programs that legacy systems handle poorly.
The industrial biotech adjacency extends the market further. Benchling customers are not only in pharmaceuticals. Corteva Agriscience uses the platform for agricultural R&D. Companies developing biofuels, biomaterials, fragrances, and food ingredients run their R&D on Benchling. Biology as a technology platform touches more industries than most people track. Each one is a potential customer segment.
The Investor Problem That Became a Moat
The original investor problem, too niche for software investors, too software-heavy for science investors, turned into something that works in Benchling’s favor over time.
Because Benchling requires both biology domain expertise and software engineering depth to build well, the barrier to entry is high. You cannot build a credible Benchling competitor by hiring engineers who don’t understand biology, and you cannot build it by hiring biologists who don’t understand modern software architecture. Both simultaneously, starting from scratch, while Benchling already has ten-plus years of product refinement, 200,000 scientists trained on the platform, and a customer base that includes most of the major names in biotech and pharma.
The network effects compound this. Each customer that adopts Benchling adds to the platform’s understanding of how biology workflows actually operate at scale. The data that comes through the system, structured and standardized by Benchling’s schema, makes the platform progressively more useful. Scientists who were trained in academic labs using Benchling’s free tier form a talent pool that flows into commercial customers, creating continuous organic adoption pressure.
AstraZeneca reported reducing DNA synthesis costs by up to 90% using Benchling. Sanofi reported saving each scientist nearly a full day per week while doubling collaboration speed. These are not marketing claims. They are the ROI numbers that enterprise buyers use to justify the license cost and the implementation effort. When the efficiency gains are that clear, the renewal rate is high and the expansion within the account is natural.
What Comes Next
The AI integration question is one that Benchling has been building toward for longer than most. The platform already contains structured biological data at a scale that few organizations have. Experiment records, sequence designs, sample lineages, workflow outputs, all captured in a consistent schema across thousands of research organizations globally.
The partnership with Sanofi in 2024 to build a digital data foundation for AI-driven R&D reflects where the platform is heading. Not AI as a feature bolted on. AI as the analytical layer sitting on top of a structured, comprehensive, trustworthy data foundation that Benchling has spent twelve years building. Predictive modeling for experiment design. Natural language processing to automate data entry and extraction from unstructured notes. Computer vision to analyze images from biological assays.
The framing Sajith has used is that biology is going through the same transformation computing went through with the internet. The tools that powered the old biology are not the tools that will power the new one. The new one is data-intensive, collaborative, globally distributed, and increasingly automated. The platform that becomes the system of record for that transition is positioned to capture an enormous amount of value.
The IPO timeline has been discussed publicly and delayed publicly as market conditions shifted. Revenue crossed $210M in ARR in mid-2024 growing 27% year over year, with the customer base growing from 410 in 2020 to over 1,200 by 2024. The trajectory is there. The question is timing.
What the Benchling Success Story Is Really About
Software really did fail science. For decades, the tools supporting biology research were inadequate for the work being done. Legacy ELN products built in the early 2000s. Excel spreadsheets stretched far beyond their design specs. Paper notebooks in the age of cloud computing.
Sajith identified this at 22, in biology labs that were simultaneously doing extraordinary science and using embarrassingly primitive software. He found a co-founder who matched his technical ability. He went through YC when most investors didn’t understand what they were building. He gave the product away to academics for years before it generated meaningful commercial revenue. He drove from lab to lab at Cal and Stanford getting rejections and feedback until the product was right.
The academic moat compounded for years and then paid off in a way that money could not easily attack. Scientists who loved the product as users brought it into organizations as employees. The pipeline from graduate program to pharma company became a sales channel that no competitors’ sales team could replicate.
The biology thesis was correct. The software infrastructure layer for the biotech revolution was genuinely missing and genuinely important. The company that got there first, with the right product and the right distribution motion, was positioned to own the category the way GitHub owns developer source control.
A kid who stayed after school in the computer lab. A leave of absence from MIT. A product that investors thought was too narrow and too strange.
Two hundred thousand scientists across 7,500 research institutions using it to build the next generation of medicines, food, and materials.
Leave a Reply