
GapQuery
Find your next micro-SaaS idea using real marketplace data
Between projects I always hit the same wall: what next? The usual answers are opinion. Pick a niche you love. Look for problems. Sure, but passion doesn't pay rent, and "look for problems" never tells you which ones. So I stopped reading idea threads and scraped the marketplaces instead: 35,600 apps across 11 ecosystems, plus their reviews and pricing.
A few things surprised me.
Bad ratings on high-review apps are the clearest signal there is. On Shopify, 13 apps have 100+ reviews and sit under 3.5 stars. Each one is a product brief with a paying customer base already installed. Even first-party isn't safe: Shopify's own Bundles app is at 2.70.
Crowded is not the same as solved. Shipping and logistics has 3,529 apps, one of the most packed categories in the store. It also averages 4.03 stars, below the 4.45 ecosystem mark. A category can be jammed and still be full of merchants who tolerate what they use instead of loving it.
Price and quality often move in opposite directions. Inventory and supply-chain apps average $193 a month at a 4.12 rating. Merchants pay double the ecosystem average for a below-average product. That gap is the opening.
And whole integration lanes sit empty. QuickBooks has 58 connectors total. Airtable, Monday, and ClickUp have almost none. The bridge nobody built is often a cleaner wedge than the tenth app in a crowded category.
Here's the pattern that changed how I pick projects: the useful question isn't "what could I build," it's "what's already failing its paying customers." Opinion can't answer that. The data can.
A methodology note: these numbers come from GapQuery, the tool I built to aggregate this across 11 ecosystems so I'd stop guessing. You can rebuild the analysis by hand from the app stores. It just takes weeks instead of minutes.
If you're between projects, tell me which ecosystem you're eyeing and I'll pull the underserved categories for it. Curious what surprises you.
About
The problem was never "what could I build." It's "what's already failing its paying customers." So I scraped 35,600 apps and their reviews to answer it with data instead of another idea list.

Comment