Do not index
Do not index
This morning, I stumbled upon an interesting product: BrowserCopilot.ai. It’s a neat idea—use page content in your browser to trigger AI for tasks like summarizing a page or replying to an email. Cool, right? But here's the kicker: I replicated it in five minutes using Cursor, without writing a single line of code.
So what’s the lesson here? Don’t use AI as the core of your SaaS product.
I’ve been saying this for years, and today’s experience only reinforces it. Here’s why:
1. The Barrier to Entry Is Too Low
If I can rebuild your product in minutes, what’s stopping everyone else? The accessibility of AI tools is eroding the exclusivity that once came with technical expertise. If your product can be copied this easily, you’re in for a rough ride.
2. Big Companies Will Catch Up Quickly
Your "innovative" AI feature? It’s probably going to be native in Chrome, Firefox, or some other mainstream tool in no time. When that happens, they’ll take all the market share. In BrowserCopilot’s case, it’s only a matter of time before browser developers bake something similar right into the browser itself.
All of this, probably begs the question: what should someone work on?
Well, imo the real opportunity lies in solving complex, niche problems for your customers.
AI can help with this, but it can’t be the entire product. It should work quietly in the background, powering solutions to real challenges—not just be a shiny wrapper for a GPT.
Given how fast you can spin up new products today, the question becomes: how do you build a defensible moat around your SaaS? For me, the answer lies in two main areas:
- Distribution Is Key
Spitting out products (or code) is easier than ever, but marketing still isn’t. Distribution remains a bottleneck, which is why I always look for integration partners—people or companies with existing audiences. Offering them a SaaS product unlocks recurring revenue for their one-time customers, which aligns our incentives perfectly.
- TTV & Lock-In
- Time to Value (TTV): How quickly can your product deliver value? The faster, the better.
- Lock-In Period: Make it harder for customers to leave, either financially (e.g., setup fees) or through data dependency. Think of a to-do app—once your tasks are logged, switching to a new one is a hassle.
AI is powerful, but it’s not the destination—it’s a tool. The real challenge (and opportunity) lies in creating solutions that your customers can’t easily replace. And that’s where the fun begins.
