You Should Probably Build

The overwhelming theme from my recent discussions with business leaders is that Software as a Service (SaaS) is in serious trouble. The Buy vs Build equation is broken and needs to be rewritten. For Small and Medium Businesses (SMBs), that rewrite is simple: You Should Probably Build.

When the head of Claude Code, Boris Cherny, says that “coding is largely solved” we should all be thinking about what that means to our business. (Check out Lenny’s Podcast with Boris if you haven’t already). Something is pretty broken when 80% of features in the average SaaS product go unused (Pendo), yet the average company still spends $4,830 per employee per year on SaaS (Zylo). It’s hard to justify a 150-feature, per-seat-app when agentic coding has all but removed the barrier to building smaller, tailored tools internally.

We’re seeing this firsthand. As a boutique consultancy, we’ve put down the generic, expensive SaaS solutions and picked up the pitchforks. If coding is largely solved, the competitive advantage shifts from who buys the best tools or has the most engineers to who builds the best workflows. Instead of off-the-shelf SaaS solutions with feature and cost bloat, we develop purpose built tools that enforce focus and create competitive advantages. The leaders I’m talking to want the same thing and know the critical workflows to optimize but struggle with where to start and are worried about security. This is where LightSight likes to partner and co-create solutions.

Because the equation has changed to “You Should Probably Build”, the real question becomes how you prioritize and manage those solutions. I recommend starting with automating the repetitive and mundane so your teams can focus on the big picture. Start where the friction is highest and the judgment is lowest. If a task is repetitive, rule-based, and your team groans about it, that’s your first build. This is tried-and-true product management: talk with users, build the thing, test, improve, test.

I realize that for many, “build” sounds like a multi-year, multi-million dollar commitment. It isn’t anymore and we are seeing that with our clients. Next, I want to spend time discussing practical business applications, including how to prioritize solutions and how to decide which SaaS seats to cut first, and crucially, what still makes sense to “buy”. My rule of thumb: Keep AI in the mundane (to automate) and tactical (to optimize) so your team can focus on strategy.

I’ve never been more excited about tech innovation, the opportunity feels monumental.

-Will