The Death of SaaS is Greatly Exaggerated

There's a popular idea making the rounds in tech circles: SaaS is a dead business model because people can now "vibe code" their own software. AI-powered coding assistants like Cursor and v0 are so powerful, the thinking goes, that every company will just build custom versions of Salesforce, HubSpot, or whatever other SaaS products they're using.

I think that's wrong. It's a little like saying people are going to 3D print their own furniture or silverware.

Don't get me wrong - vibe coding tools are genuinely powerful and transformative. They're changing how we build software in fundamental ways. But the idea that they'll replace SaaS fundamentally misunderstands both how software development actually works and where these tools add the most value.

Here's why the "death of SaaS" narrative is overblown, based on my experience building Quotient and maintaining software at scale.

The 90% vs 1% Illusion

Getting to a serviceable prototype for a new CRM is easy with vibe coding tools, and it feels like you're 90% of the way there. But actually you're only 1% of the way there.

This is the most insidious part of the "we'll just build it ourselves" mentality. Because the initial build is so fast and feels so productive, it creates a false sense that building production software is easy now. You spin up a working prototype in an afternoon and think "wow, we're basically done!"

But the real work of software development isn't building the prototype. It's everything that comes after:

• Fixing bugs that only appear in production with real users

• Responding to user requests and feature additions

• Upgrading packages when security vulnerabilities are discovered

• Maintaining third-party integrations as APIs change

• Scaling infrastructure as usage grows

• Training new team members on a custom codebase

These aren't edge cases or corner scenarios - this is the actual substance of software development. I can personally attest that this is the stuff that takes upwards of 90% of our team’s time, and most of it isn’t sexy or fun.

The prototype is the easy part. Vibe coding tools make building the initial version faster, but they don't fundamentally change the ongoing maintenance burden. And that burden is enormous.

SaaS Companies Have the Same Tools

People seem to forget: SaaS companies also have access to vibe coding tools.

So in theory they should be able to leverage them to make their SaaS applications much bigger and better. We've already seen this at Quotient - these tools have allowed us to be much more ambitious with the product we've built. We are able to build more and higher quality features with a smaller team thanks to AI-assisted programming.

Vibe coding tools or not, your little internal team still has vastly less manpower than Salesforce's or Adobe's engineering team. They have hundreds or thousands of engineers who can now use these same tools to ship features faster, fix bugs more efficiently, and maintain more sophisticated integrations. You don't have some unique advantage over the existing SaaS vendors. They have these tools too. Theoretically they should use them to improve the quality of their solutions - and the smart ones are doing exactly that.

The playing field isn't being leveled by AI coding tools. If anything, it's tilting further toward companies that already have strong engineering teams and deep domain expertise in their problem space.

Not Your Core Competency

This is the most important point: at the end of the day these SaaS applications simply are not your company's core competency.

Companies will quickly realize that their time and energy is better spent adding value to their own products and services, rather than trying to save a few bucks on their CRM bill by creating a Salesforce clone. To the extent that companies adopt vibe coding, they'll be better off using it to improve their own products and services.

The fintech company building revolutionary payment infrastructure will focus on making their payment APIs faster, more reliable, and more feature-rich - not rebuilding Salesforce. The SaaS startup will invest engineering cycles into their core platform rather than custom CRM development.

The opportunity cost is staggering. Every hour your engineers spend maintaining your custom-built CRM is an hour they're not spending on your actual product. And unlike Salesforce, your internal CRM isn't generating revenue - it's just overhead.

So where does vibe coding actually fit? It's not for replacing your core business software stack - but there are plenty of places where it makes perfect sense.

Where Vibe Coding Actually Makes Sense

All that being said, I still absolutely think there is room for vibe coding inside large organizations. But it's more to fill in the gaps between existing software solutions or to build bespoke applications that are specific to your business.

Every business has tons of internal tools, Python scripts, and data visualization dashboards that are slapped together, often by non-engineers. I think that vibe coding has the potential to make these much easier to build, and I'm very excited about that.

At Quotient, we vibe code:

• Internal tools and dashboards - I think vibe coding is in particular a great replacement for traditional BI tools like Looker and Tableau. Need a custom dashboard to track specific metrics? Build it in an afternoon instead of fighting with a rigid BI tool.

• Python and bash scripts - One-off administrative tasks, like pulling specific lists of customers or crunching usage numbers. These are perfect for vibe coding because they're small, focused, and don't need long-term maintenance.

• Internal wikis and documentation - Custom documentation sites for our codebase that integrate with our specific workflow and tools.

These are perfect use cases for vibe coding because they're:

• Specific to your business and unlikely to be well-served by generic SaaS products

• Simple enough that maintenance burden is minimal

• Filling gaps between your existing software solutions

• Often built and maintained by the people who actually use them

This is very different from trying to rebuild Salesforce or HubSpot. You're not replacing core business software - you're augmenting it with custom tools that solve your specific problems.

The Reality Behind the Hype

Like so many things in tech, there is some reality behind the hype, but a lot of it is quite overblown.

Vibe coding tools are genuinely transformative - they're making developers more productive and enabling teams to build more ambitious products. But they don't change the fundamental economics of software maintenance or the value proposition of specialized, well-maintained SaaS products.

If anything, SaaS is going to become more valuable as it becomes imbued with AI agents that can do increasingly complex tasks for you. The future isn't everyone building their own CRM - it's using increasingly intelligent SaaS products that adapt to your needs while someone else handles the maintenance burden.

So use vibe coding for what it's great at: filling the gaps, building internal tools, and improving your actual product. But don't fall for the idea that you're going to replace your entire SaaS stack by building everything yourself. That's not how this plays out.