Micro Apps - the re-birth of SaaS

I've been thinking about what it would look like to find an app and have it optimized for my use case, instead of choosing a SaaS tool with all these random features I don't need. I think AI could make that a different experience. One of my first thoughts was a model like YouTube, where it would allow smaller creators. But instead of content creators, I think there's going to be a world where the people that know how to use AI can make two to five apps a day.

They would be creators in the sense that, instead of paying $50 a month for QuickBooks and all these random features I don't need, I could basically pay $200 and have somebody make an app for me. Or customize something they already have for my particular use case. I could find the one that I like, see what people are liking, and sign up to have it optimized for my use case. I wonder if maybe that's more along the lines of how it could play out. You'd have creators creating highly relevant, high context apps for your use case.

For example, someone who needs a market newsletter might say, "I need something that tells me the market data once a week and sets up a newsletter and sends it out." There are 50 newsletter apps, but this one has a lot of reviews. Everybody loves it. They submit all their requirements, then starts working with the creator's virtual expert to build the system with some of the proprietary plugins he's created. He could hop in and validate it, or they could decide it's already working the way they need.

For me, I'd be able to specify what I need with our business. We have four different bank accounts and three different properties. Each property has capex and rentex categories, and we like to do travel hacking for our personal finances. Every once in a while, things need to be put in manually, or the wrong card with the wrong account gets charged and needs to go to the other business. I need to track the receipts. I feel like AI could put together a plan. If somebody connected Plaid and an S3 bucket and all these things, I could see what it would cost to process my receipts with the product that guy has made.

Basically, an AI-powered Product Hunt. You see which ones people love, and you could augment them with skins other people have made for the core functionality. That kind of an experience, I think, makes a lot of sense. Some people might not need a bunch of different entities. They might just want to track their Uber stuff, and they already have their personal finances figured out somewhere else. Somebody else might want a newsletter, an email service, or a contract and project tracking service. I'm doing contracts over here and FreshBooks over there, and then this person needs me to manually input over here. My needs for our tracking are highly complex. Finding four, five, or six different SaaS tools to do what I need is a really high effort, high burden process.

I think one of the things AI is going to possibly do is make it really easy to find and customize tooling into one interface for the different types of things you want. Real estate investing, client management, graphic design, video creation, content editing. Those kinds of things could all roll up into your AI interface. Everybody would maybe have one app that they go to do everything, and then have all of the services and the agents rolling up to that one interface. I feel like that's the future people would like and want. If I need it to do something new, it can augment what I'm already familiar with instead of having to switch tools completely.

If I consume different things in a way that's familiar to me, then I can upload them to some sort of system that can map them to what's familiar to someone else. My Trello board and your Trello board might be the same Trello board, but how I get information out and enter information could be highly optimized to me. Maybe a company subscribes to Trello to make it really easy for people who don't have AIs to still come in and interface with things. But someone like me could have a highly customized way to interface with that service.

So SaaS isn't necessarily dead. I think maybe it's infinitely more customizable. Maybe that's a better way to think about it.