At the current moment, “Shadow AI” and “SaaS is dead” are two popular topics. I realized that there was an interesting intersection here. With vibe coding, anyone can replicate a web product’s functionality in the weekend (or so the theory goes). The reality is a little more complicated

If you build it, you must support it Link to heading

If you application has users, you need the application to do things for those users. Even if the only user is you. If no one is using the application, it might as well not exist.

That maintenance might be cheaper, but it still has cost. At minimum, you have to shift focus to have an agent start a fix, and evaluate if the fix works. If you have users other than you, they have to communicate with you, and you have to communicate with them.

All of this consumes time and focus that could otherwise be applied to making the revenue necessary for your company to succeed.

Once users are engaged with a product, they’ll find things missing. Pareto Principle (80/20) always applies, as does Zeno’s Paradox. THere’s alsways infinite space between where you are and the “perfect” version of the product. This gap historically is part of what drives revenue for SaaS companies - you productize things in the gap, and sell them to customers.

When you’re vibe coding, you don’t have paying customers - just internal ones. So you have a growing set of uncompensated features you’re developing. This compounds the issue above, as all forward feature work becomes a 100x maintenance burden.

Shadow Engineering enters the mix Link to heading

So what is an aspirational vibe coder or founder to do? History teaches us what tends to happen: the feature work and maintenance gets done, but gets pushed below the fold. You work on core priorities during the day, and swap to this work during nights and weekends. In many cases, you ignore your core priorities or compromise on their quality.

This isn’t limited to founders - many production products are partially supported by invisible work that flows outside the existing process. It’s a natural human thing to do, especially when corporate and personal priorities fail to align.

Enterprise companies can afford this misalignment though. Startups cannot. If a startup devotes a month of engineering-time to an off-topic initiative, that’s a month (or more) that they’ve fallen behind on product market fit.

Allowing or incentivizing shadow engineering within your organization is a great way to ensure you don’t have an engineering organization next year.