How to Find and Rebuild Underperforming Apps with AI

Kyle Balmer breaks down a playbook from Greg Eisenberg for building profitable AI-powered apps. The strategy: find apps that people pay for but dislike, then rebuild them with AI doing the heavy lifting.

Step 1: Find Apps People Pay For but Hate Using

Head to the App Store, Google Play, or a tool like Sensor Tower. Look for apps that generate real revenue but have four-star or lower reviews. This tells you there is market demand, but the existing products leave customers unsatisfied.

Greg Eisenberg's post outlining the playbook for rebuilding existing apps

The goal is to identify where money is flowing in but users are frustrated. That gap between willingness to pay and product satisfaction is your opportunity.

Step 2: Spot the AI Opportunity

Once you have identified a target app, figure out whether AI can replace the core functionality. Look for apps where the main feature is some kind of automation or data processing that AI can now handle faster and better.

Eisenberg's post detailing how to spot AI opportunities in existing apps

Meal planners are a good example. Users enter their information and the app generates a plan. That used to require complex programming. Now an LLM can do it instantly, and the output can be more personalized.

Step 3: Move Fast While the Big Players Move Slow

You might think that apps like MyFitnessPal or Duolingo will just add AI features themselves. They are already experimenting. But large companies are slow. They have legacy codebases, internal politics, and long release cycles.

Discussion of how large apps will eventually add AI features but move slowly

That slowness creates a window. You can build something lighter, cheaper, and faster. From there, you have two paths: carve out a niche and generate income from it, or build enough traction that one of the big players acquires you outright.

Key Takeaways

Look for apps with strong revenue but poor reviews to find rebuilding opportunities.
Target functionality that AI can replicate or improve, especially manual automations.
Speed is the advantage. Large incumbents will add AI features eventually, but they move slowly enough to give small builders a real window.

Resources

Sensor Tower -- App intelligence platform for finding revenue and review data
Greg Eisenberg -- Source of the original playbook shared in the video