SaaS vs In House & Open Source

TLDR: Generative AI has changed the calculus around “build vs. buy.” As the cost of software development goes down, it becomes far more feasible to solve problems in-house—raising the bar for when SaaS solutions make sense.

When software was difficult and expensive to create, businesses could charge tens of thousands of dollars just to build a simple website or blog. Over time, as web development became more accessible, more people and organizations gained the ability to build and operate these websites themselves. We’re now seeing the same shift for more complex web applications—those with rich user interfaces, workflows, and data management needs. Infrastructure is still challenging, but the trend is toward greater accessibility, which changes the economics of “let us host it for you.”

That doesn’t mean SaaS is going away. Applications that solve broadly shared problems with huge market sizes—like Payroll and HR—are well suited to SaaS. It rarely makes sense for a biotech organization to build or maintain its own HR system. But many problems in biotech aren’t that general. Even relatively “general-purpose” categories like LIMS and ELNs often require heavy customization before they’re a good fit within an organization.

For more bespoke problems, the calculus has shifted most dramatically. In the past, it might have made sense to lean on a niche SaaS provider for a specific workflow or data task. Today, it’s often faster and more practical to build something internally using the growing ecosystem of open source tools, cloud platforms, and generative AI–powered development resources. These bespoke challenges are often where biotech companies create their most unique value—because the point of biotech is to work on problems others aren’t tackling.

Data sensitivity adds another layer of complexity. Intellectual property is the core of most biotech companies, and regulations around data security and export controls are only becoming stricter. That makes it essential to carefully evaluate where SaaS fits, and to have strong in-house capabilities for anything touching proprietary data.

Of course, this doesn’t mean you should always build in-house. For early-stage biotechs, leaning heavily on SaaS is usually the right approach—it allows teams to move quickly and stay focused on science. As companies grow, though, the balance naturally shifts. The most successful organizations will be those that plan ahead to build a strong in-house computational team, capable of combining SaaS, open source, and custom solutions into a strategy tailored to their unique needs.