The Two Markets for Software
Stuck in Bangalore traffic for two hours, I found myself looking at the metro bridge under construction. I imagine a future where the construction is complete and think about the countless man-hours it will save—not just in commercial value, but in reduced frustration, noise, and inefficiencies. It reminds me of a scene from Margin Call, where a character explains that a bridge he designed saved 1,531 human years.
Later that week, my car tire got punctured—twice. At the petrol pump, I asked for a rope plug, a cheap and effective fix. The attendant insisted on an internal patch, a solution three or four times as costly. I didn’t budge. I chose the cheapest solution that got the job done.
Billions are invested in infrastructure like metro bridges because we recognize their long-term value. But when it comes to software—one of the few levers that can deliver massive productivity gains—we treat it like a tire puncture fix. Instead of investing in scalable, reusable, high-quality software, companies settle for cheap, short-term patches that ‘just work’ for now.
Software is sold in two very different markets:
1️⃣ Companies that understand software’s leverage—how automation and scale can 10X productivity. They know that software working 99% of the time is exponentially more valuable than one that works 90% of the time.
2️⃣ Companies that see software as a commodity—‘Just give me the cheapest thing that works for now.’
I've spoken with founders and executives of large companies that purchase software, and nearly all of them are disappointed with the quality. The biggest frustration? It's not reusable. Many legacy IT service providers build archaic, non-scalable code that constantly breaks—creating a cycle of dependency that ends up costing far more for maintenance in the long run than simply investing in high-quality software from the start.
If software keeps being treated like a disposable commodity, the incentive to build high-quality solutions disappears? This mindset keeps us trapped in a cycle of poor software, constant replacements, and endless frustration
We're treating software like a cheap tire patch—good enough to stop the bleeding for now—when we should be building bridges that last.