557,000 new apps hit the App Store last year
A 24% jump. The biggest wave since 2016. December alone was up 60% year-on-year.
And here's the kicker, Apple is now cracking down on vibe coding apps like Replit because the flood is literally breaking their review process.
Everyone can build software now. Which means software isn't the advantage anymore.
I say this as someone who's done it. I've vibe-coded three products in the last few months, ReviewFlow, NourishRx, Equiptech. No engineering team. No co-founder with a CS degree. Just me, Claude, and a weekend.
It's genuinely incredible. The barrier to building software has collapsed to zero.
But that's exactly the problem.
When everyone can build, building isn't the moat.
The SaaS market is getting absolutely wrecked right now. The software ETF is down 21% this year. $2 trillion in market cap… gone. And for the first time in modern history, SaaS stocks are trading at a discount to the S&P 500.
That's never happened before. Ever.
There are now 30,800+ SaaS companies worldwide. The average business uses 106 different SaaS apps. And every single day, someone on Twitter is announcing their new AI-powered [insert literally anything] tool.
AI invoice tracker. AI habit coach. AI meal planner. AI bookmark manager. AI meeting summariser and i could go on for literally the next 5 mins, as you all know. i mean a few of you reading this now just put their head in the hands releasing this right now.
Every niche is about to have 100 mediocre versions of the same app.
So what actually wins?
I keep coming back to three things.
1. Distribution beats product
The best app in the world dies in obscurity without distribution. The mediocre app with a killer acquisition channel prints money. This isn't new, but it matters more now than ever because the product gap between competitors is shrinking to zero. If anyone can build it in a weekend, your edge is never the build. It's how you get it in front of people.
Shopify's CEO told his entire team to prove AI can't do a job before they're allowed to hire a human. He cut headcount from 11,600 to 8,100 while growing 20-40% a year. He gets it. The leverage is in the system, not the headcount.
2. Physical presence is a moat
I run restaurant locations. Nobody is vibe-coding a burrito i mean apart from Travis Kalanick. Nobody is prompting their way through a Friday night rush. The businesses that require a human body to show up, a physical location to exist, those have a natural defence against the wave of AI-built competitors flooding every digital niche.
The irony is wild. Tech people spent a decade telling everyone that software would eat the world. Now software is eating itself, and the businesses that require atoms instead of bits suddenly look like the smart bet.
3. Labour-intensive = defensible.
Trades. Cleaning. Construction. Aged care. Logistics. These aren't sexy. They don't get Product Hunt launches or Twitter threads with rocket emojis. But they have something that no vibe-coded SaaS ever will, they're hard to replicate because they require people, training, relationships, and showing up.
The best play I can see right now? Use AI to run a labour or physical business better, not to build another app competing with 100 identical apps.
Here's what I'm actually doing about this.
I'm not stopping building software. But I'm thinking about where software creates leverage versus where it's just another commodity.
ColdPulse (aircon monitoring agent) works because it's tied to actual hardware on actual roofs. NourishRx is the one going to the graveyard, it's pure digital, pure SaaS, competing in a space that's getting crowded fast.
The pattern is clear. The closer your product is to the physical world, the wider your moat.
Everyone's building apps right now. The people who win won't be the best builders.
They'll be the ones who own the distribution. Or the ones whose business requires a human to show up.
Probably both.