Buy vs Build

Is generative AI, and vibe coding in particular, ringing in the end of SaaS (Software-as-a-Service) as we know it? After all, why pay monthly for expensive software when you can code it up for yourself?

Tech Influencers have you believe that you'll save thousands, and list all the things they've coded up in a weekend.

I'm skeptical.

Time Versus Money

Even before generative AI, any company had to decide whether to buy certain capabilities or build them in-house. Generative AI has shifted the costs of building somewhat, but has certainly not eliminated them. One of the benefits of buying a solution is not just that you save the upfront build cost. You also save the ongoing maintenance cost.

More than anything, a company wants to focus all their energy on the unique value it brings. Unless your company's product is a calendar booking app, it's a distraction to vibe-code your alternative to Calendly rather than just paying the subscription cost.

Where in-house vibe coding has its place are small tools for which, before Gen AI, you would never have thought of getting custom software made for it, because the return on investment would be too low. If it costs tens of thousands of dollars to build a tool that, realistically, only saves you a thousand bucks each year, such a project will not happen. If generative AI drops that cost down to a few hundred dollars, real savings can be had. Though, in the grand scheme of things, they'd still be small.

Instead of chasing such small gains, it's better to focus on putting AI to use where it can drastically enhance your unique value proposition. What is it that only you can do, and how can AI give you a boost there?