Overview of Rumors of the SaaS demise are greatly exaggerated
In this episode of the Stack Overflow Podcast, host Ryan Donovan speaks with engineering leader Nivea Henry about whether AI is accelerating the decline of SaaS or simply changing how teams evaluate build vs. buy vs. borrow. The conversation argues that SaaS is not dead, but that many software decisions are being reshaped by AI, higher customization expectations, security concerns, and the rising ability for engineering teams to build internal tools quickly.
Key Discussion Points
AI is changing the build-vs-buy calculus
- AI makes it easier for engineering teams to say, “we can build this in a couple of weeks.”
- That shifts the old default toward buying SaaS, especially for:
- simple workflow tools
- internal productivity software
- highly customized team-specific processes
- But the real question is no longer just “Can we build it?” It’s “Who maintains it, secures it, upgrades it, and supports it?”
SaaS still makes sense in the right situations
Nivea emphasizes that SaaS remains valuable when:
- the workflow is complex
- the organization lacks bandwidth
- the software requires ongoing operational maintenance
- the use case benefits from vendor-managed security, compliance, and support
For certain categories, especially infrastructure-heavy or broadly used tools, SaaS still offers real advantages.
Open source is still useful, but under pressure
- Open source remains a strong middle ground between building and buying.
- However, the episode highlights growing concerns around:
- maintainer burnout
- “slop” contributions
- abandoned projects
- companies leveraging community labor and then privatizing or walking away
- Nivea warns that when open source ecosystems weaken, the loss is not just technical—it also reduces shared learning and community knowledge.
Internal tools may be safer than third-party tools
A major theme is that for some use cases, building internally can actually be more secure and less risky than outsourcing:
- Less exposure to third-party access
- Fewer compliance and privacy complications
- Reduced risk of prompt injection or external data leakage
- More control over how workflows evolve
This can make internal development the better choice even when a SaaS option looks cheaper upfront.
Microservices and modular tooling are under renewed scrutiny
- Nivea takes a skeptical view of the long-term promise of microservices.
- She argues that many companies adopted microservices because they mapped neatly to org structure, but the overhead became massive.
- Her recommendation:
- start with a monolith when appropriate
- only split out the services that truly need separation
- avoid overengineering unless scale and complexity justify it
The real moats are data and cost control
According to Nivea, the strongest defenses for SaaS companies are:
- Data: owning unique insights, analytics, and value-added intelligence
- Cost management: controlling variable infrastructure costs better than customers can internally
She argues that SaaS vendors need more than just access to your data—they need to offer something you can’t easily replicate yourself.
What Engineering Leaders Should Think About
Use systems thinking, not just code thinking
Nivea stresses that engineers should understand:
- the business context upstream of development
- how work enters the system
- how tooling affects the broader software delivery lifecycle
- where data flows and fragments across systems
Her point: if you only optimize the code layer, you may miss the bigger operational and organizational consequences.
Watch for tool sprawl and data fragmentation
Too many third-party tools can create:
- fragmented data
- inconsistent workflows
- weaker observability
- lower-quality insights
- more operational complexity
Even if a tool seems convenient, it may create long-term system problems.
Ask better questions before choosing a tool
Nivea repeatedly returns to three core considerations:
- Cost of licensing
- Cost of ownership
- Cost of growth
Those three questions help determine whether to build, buy, or borrow.
Notable Takeaways
- SaaS isn’t dying so much as becoming easier to challenge.
- AI lowers the barrier to internal tool-building, especially for repetitive or customizable workflows.
- The best SaaS products will likely be those with real moats: data, analytics, compliance, and hard-to-replicate infrastructure.
- Open source is valuable but fragile if maintenance and community support erode.
- Engineering teams should think upstream about business processes, not just downstream about implementation.
- Not every tool needs to be bought—some can and should be built internally when the use case is narrow, repetitive, or highly sensitive.
Bottom Line
The episode’s core message is that the “SaaS apocalypse” is overstated. What’s really happening is a more disciplined, AI-influenced re-evaluation of software decisions. For many teams, the answer is no longer automatically “buy SaaS”—it may be build internally for speed, control, and security, especially when the tool is simple or highly specific. But for vendors to survive, they’ll need to offer more than functionality: they’ll need to deliver insights, trust, and durable value.