The Statue of Liberty was gifted to the United States by France in 1885 to celebrate the first centennial, the continuation of American democracy, and the recent abolition of slavery. It remains a powerful symbol for hope and liberty 140 years later, but it almost never got built.

That's because those French pricks only shipped over the statue. It was up to New Yorkers to construct the foundation. No base, no statue.

A lack of funding was the biggest problem. The United States just exited the Long Depression, which spanned 1873 to 1879, and was still rebuilding from the destruction of the Civil War.

In a last-ditch effort, Joseph Pulitzer launched a crowdfunding campaign in his New York World paper. Emma Lazarus wrote the poem "The New Colossus" to drum up support, which played a role in 120,000 individuals donating to the cause. Now immortalized on a bronze plaque on the statue's crowdfunded-foundation, the poem gave us the infamous lines:

Give me your tired, your poor,
Your huddled masses yearning to breathe free.

This is similar to how technological hype works. The narrative is always cool, but outcomes are what matter.

It's awesome to dream of a future where computational tools greatly increase the speed and accuracy of developing drugs (the statue), but we won't get very far if we never build a solid foundation. Science takes time and is often uneventful, much like crossing the ocean on a steamship.

If scientists don't see the promise of new tools as they near the harbor (the huddled masses), then all the hype might never translate into real-world discoveries (breathe free). We'd just have a story about what would've been a cool statue – and one less reason to like the French.

Certara has largely missed the AI hype train. Its customer base has been cutting costs for four years and it still generates more revenue from lower-margin services than high-margin software. But investors are forgetting that services are the foundation for future software growth. Not just software-as-a-service, but actual services with human beings and endless Zoom meetings and site visits involved.

The recent launch of Certara IQ, the world's first integrated software platform in the accelerating field of quantitative systems pharmacology (QSP), could eventually enable double-digit and high-margin growth for years to come.

What is Quantitative Systems Pharmacology?

Let's back up for a second.

Part of the reason SaaSpocalypse fears are gripping Certara is because many analysts and investors don't understand how research is conducted in the real world.

Drug development requires the use of dozens of individual tools, including both computational modeling tools (virtual) and experimental tools in the lab (physical). The outputs from one tool become the inputs for another. It's sometimes possible to build research workflows as iterative loops, called the design, build, test, learn (DBTL) cycle, where systems spit out improving results over time. The first iteration (n=1) might not be very close to answering a research hypothesis, but the thousandth (n=1,000) pass might yield more conclusive results.

In an ideal world, these input-output handoffs could be automated with robotics and software, enabling semi-autonomous labs capable of testing thousands of experimental designs every day or week. Then nerds in lab coats could stop moving tiny volumes of liquids around all day and instead sit back and use that hulking grey matter to choose the best development candidates to move forward.

In the real world, it's still difficult to automate many R&D workflows, but higher-throughput systems are becoming more widely used. That spans large research centers like AstraZeneca's DISC and even smaller drug developers oriented to a data-first approach.

‍

QSP is an integral drug discovery and drug development tool

As experimental throughput increases throughout the industry, the volume of data generated from each iterative pass increases exponentially. The best way to make sense of it is to lean heavily into Model-Informed Drug Development (MIDD) tools, such as QSP.

QSP is a biosimulation approach that unifies computational and experimental data to inform drug development. It can be used in both drug discovery and drug development activities, although few drug developers collect enough data from clinical trials for it to be very useful for the latter. The largest pharma companies are routine exceptions. Relay Therapeutics, the first QSP-native drug developer, is another. It's the reason the company could confidently alter the recommended phase 3 dose (RP3D) for zovegalisib in the ReDiscover-2 study; a decision largely misinterpreted by investors.

Importantly, computational modeling must be confirmed and validated with experimental datasets. A drug developer might use QSP as the first step in drug discovery when experimental data aren't available. As a program matures, data from physical experiments are generated, and QSP models evolve and adapt to that specific program.

Drug developers cannot just ask a machine or algorithm how a never-before-seen chemical entity will behave and get an intelligent answer. The most valuable applications come from purpose-built models. These are very complex. Deploying these tools will be different for an antiviral vs. an antibody, or for a solid tumor vs. a liquid tumor, or for a double combination therapy vs. a triple combination therapy.

What are some real-world applications?

Pfizer used QSP to accelerate the development of Paxlovid (nirmatrelvir/ritonavir) during the coronavirus pandemic. It modeled the unique viral replication characteristics of coronaviruses, how the immune system responds, and how that system changes when specific antivirals were introduced.

The company had initially proposed a 10-day treatment window, but purpose-built QSP modeling led researchers to determine that maximum efficacy might be achieved in only five days. As a result, Pfizer was able to design a shorter study with fewer patients. Halving the treatment window also reduced healthcare costs and inconvenience to patients, while doubling the return on investment for Pfizer's manufacturing infrastructure.

These aren't hypotheticals for investor story time. These are real-world dollars and cost efficiencies.

You might be surprised to learn that QSP isn't new. The first FDA submission to utilize the approach was submitted in 2013 for the thyroid medication Natpara from Takeda.

However, it took time for regulatory frameworks to catch up, as evidenced by the number of regulatory submissions vs. the number of publications in the scientific literature. Every submission incorporating purpose-built QSP tools requires paired meetings with regulatory scientists.

These barriers have gradually been removed as the FDA gains more experience and confidence in biosimulation data. The number of QSP-based submissions has roughly doubled every 18 months since 2013 and now represents over 5% of all filings, but the pace has been much faster in recent years.

‍

Can Certara IQ remove a key bottleneck to QSP adoption?

Regulatory frameworks aren't the only bottleneck to wider adoption. The complexity and time required to build QSP models is another hindrance. In fact, drug discovery programs incorporating the approach actually take more time than previous empirical approaches. The difference is measured months, but the delay does eat into any time savings in later development. This is often overlooked.

Enter Certara IQ.

Certara designed its newest software platform to address the current bottlenecks of QSP adoption.

• Standardization: It standardizes model building, which allows teams to re-use parts of bespoke models from unrelated programs and better understand when to add or remove components. That saves time and costs, but also greatly increases the reproducibility and accuracy of results.
• Pre-built models: Certara has leveraged its deep expertise to develop validated QSP models for specific therapeutic modalities and therapeutic areas, allowing customers to quickly start generating insights without starting from scratch.
• Low-code platform: End users no longer need to sift through complex coding environments. Some may never need to write a single line of code. These abstraction layers can quickly drive adoption for digital tools. If you had to write code just to use Microsoft Word, then I doubt many people would see the point. Instead, you hit a letter on your keyboard and it appears on your screen; you never see the code in the background.
• Regulatory-ready: Biosimulation data from Certara IQ can be exported in regulatory-ready reports. That'll remain a key advantage even if the business sells its regulatory software segment.

Whether Certara IQ matters to the business this year is impossible to know. There's only so much the company can do to overcome budget constraints from customers.

Certara might grow by low single-digits in 2026, which has implications for where the stock price is headed in the next 12 months.

But QSP will only become more indispensable as computing power improves, we get better at measuring biology, and high-throughput experimentation becomes more widely adopted. If Certara IQ delivers for customers, then it could become one of the most important assets for the business over time.

It could be enough to remove SaaSpocalypse fears and unwind some of the stock's brutal sell off, but investors simply have to wait for the upcoming full-year 2025 earnings call scheduled for Thursday, February 26 before the market opens.

News Flow & Modeling Insights

(No change from the Q3 2025 update.)

The current model assumes the following full-year 2025 operating metrics:

• Full-year 2025 revenue grows 8.3% to $417.495 million (vs. 10.7% and $426.490 million previously). Company guidance expects a midpoint of $417.5 million (vs. $420 million previously). This includes $102.305 million in Q4 2025 revenue, representing year-over-year growth of 2%.
• Services revenue increases roughly 2.3% to $234.851 million (vs. 4.5% and $239.762 million previously). This includes $56.555 million in Q4 2025 revenue, representing a year-over-year decline of 3% (vs. $61.576 million previously).
• Software revenue increases 17.3% to $182.644 million (vs. 19.9% and $186.729 million previously). This includes $45.750 million in Q4 2025 revenue, representing a year-over-year increase of 8% (vs. $50.629 million).
• The revenue mix reaches 45.1% software revenue at the end of 2025. Software revenue ended at 42.1% in 2024, 38.2% in 2023, and 33.7% in 2022.

Margin of Safety & Conviction

Certara is considered a Current Compounder position with the following Conviction rating.

• 1 = High
• 2 = Above Average
• 3 = Average
• 4 = Below Average

The estimated fair valuation based on my current model is below:

• Market close February 13: $6.62 per share
• Modeled Fair Valuation: $11.16 per share
• Allocation Range: Up to 10%

Certara reported 159.273 million shares outstanding as of November 1, 2025. The modeled fair valuation above assumes no change in the number of shares outstanding through the end of last year.

Further Reading

• November 2025 research note analyzing Q3 operating results
• October 2025 press release announcing the launch of Certara IQ