AWS built its reputation on offering raw building blocks, a model that worked when customers expected to assemble their own infrastructure. But as The Register columnist Corey Quinn points out, newer platforms such as Vercel and Railway now offer a more polished, customer-forward experience. These platforms don't carry AWS's two decades of legacy baggage, and they don't make users prove themselves before they can be productive.
Quinn argues AWS has shown institutional arrogance, assuming customers will always choose it. He notes that AWS missed OpenAI's early move to Azure and has been playing catch-up ever since. Even AWS's own attempts to build higher-level, SaaS-style products have repeatedly stalled or failed, with the exception of Amazon Connect, which is now being repurposed as a broader brand.
The column concludes that AWS should stop trying to be something it isn't. Instead of chasing SaaS success, it should embrace its role as a wholesaler: make life better for the Vercels of the world and for small startups, rather than forcing everyone to assemble everything themselves. That, Quinn argues, is where AWS's real strength lies.