Post

I think this is exactly the math that is going to drastically change the small SaaS landscape why would I pay monthly for <tool> when the same cost (maybe less!) can make that and also anything else I need

Nick Fisher@hydroxide.dev

I pay $100 a month for Xero (accounting software), and I do wonder if upgrading to Claude Code Max for a month or two would pay for itself to write a robust replacement. It probably won’t get bank feed imports right, but that’s broken in Xero for some accounts anyway.

3 reposts57 likes

I think it’ll change the landscape more in pressure on price as alternatives crop up. I’ve coded my way out of at least one expensive subscription. But you still need to be good at project management and at least get the theory of the architecture you’re building or else it’ll be fragile as hell.

I think it’s still a really long ways away from an average person asking for what they need and the system giving it back to them functional, durable and iterable. The average person doesn’t really even know how to ask for what they need.

The new thing is they can hire an intermediary to turn their requirements into a custom system at a fraction of the previous cost of doing, and maintaining, custom software.

bespoke software doesn't need to be durable and iterable, it just needs to output usable tool contexts once from a natural language prompt claude made a working DAW that got file recorded, saved an acceptable audio clip in a couple minutes more trustworthy in low stakes art/music contexts though

+1, totally there's very different markets, skills and appetites between randos like us rolling our own tools (very rare) and some median dev who is already in charge of trying to chip away at some corpo need themselves suddenly having more powers

I suspect we’ll see more bundling of tools into suits at a lower cost. Corporate environments will have less tolerance for nickel and dimed features at $19/month. Tool sets just need to be cheaper than paying an in house person to build and maintain everything themselves + security concerns.

I got so mad yesterday that someone tried to get my email before using a vibecoded widget game that I went and started to make my own instead of just like giving them some burner email

Not just SaaS but paid libraries. Why would a dev team pay $1k for a small, single-purpose library they're probably going to have to vendor and patch anyway, instead of using Claude to NIH their own?

Curious how confident you feel in this. It feels like the sort of thing that is absolutely going to change my day-to-day, but I also feel that I will be one of the only people in my circles to take advantage of this - everyone else is gonna keep paying for quickbooks.

Oof I dunno the PDF spec was one of my first document-read-jump-scares. The model that I’m leaning towards these days is more ‘carpenter building their own jigs’ than ‘carpenter building their own saws’

in the way that we see distributions now probably, like the 1-9-90 rule or similar I suspect there will still be (always be?) a market for making a thing and selling it because people can't be bothered or are unaware of rolling their own I also suspect that the ratio of that ROI will shift around

it's also a market thing - we see this already with non-software industries relying on old crappy software every day someone makes a garbage truck route scheduling program and sells decades because no one else is interested in making a sleeker one maybe vibe coding will _increase_ saas competition

What a weird example. Is no one interested, or is it because traveling salesman is an infamously impossible problem to solve, and so once a working route is set up you hardly ever need the software again?

Deployment and data/storage management will be the sticking points but platforms will probably adapt or pop up in that space. Like, your vibe-coded to-do app is great until you add a feature or fix a bug and wipe all your production data. That will matter more for some uses than others.