Why Your Software Costs Too Much

If you run a small business and pay for software, you already sense something is off. Your scheduling tool costs $12 per month. Your email marketing platform costs $100. Your CRM costs $50 per seat. You're paying hundreds of dollars a month for tools that, in most cases, are thin interfaces over commodity infrastructure.

The feeling is correct. Something is off. The cost of building and operating software has collapsed over the past few years. The prices you pay haven't.

Three forces that changed everything

1. AI-assisted development

In 2023, a competent software engineer could write maybe 200 lines of production-quality code per day. In 2026, that same engineer, working with AI tools, routinely produces 1,500-2,000. Not boilerplate. Not copy-paste. Tested, reviewed, deployed code.

This isn't a marginal improvement. It's an order-of-magnitude shift in what one person can build and maintain. A single developer can now do the work that required a team of five just three years ago. The entire Forge IDE -a desktop application with a computational engine, package manager, and graphical interface -is maintained by one person with AI assistance.

2. Commodity cloud infrastructure

A production-grade server that handles thousands of users costs about $30 per month. Object storage is fractions of a penny per gigabyte. Managed databases, CDNs, TLS certificates, container orchestration -all commodity services now, priced at commodity rates.

What required a DevOps team and $10,000/month in 2018 runs on a single VPS with automated deployment in 2026. The infrastructure layer of software has been commoditized so thoroughly that it's essentially rounding error in most business budgets.

3. Open-source maturity

The building blocks are free. Web frameworks, cryptographic libraries, database engines, PDF renderers, email delivery SDKs, payment processors -these took millions of dollars and decades of engineering to build. They're available to anyone under permissive licenses.

No one is writing an HTTP server from scratch anymore. The leverage from open-source has been compounding for thirty years, and we've reached the point where most of what your SaaS tool does is orchestration of existing open-source components.

So what does software actually cost to run?

Let's do the math on a scheduling tool. The kind that lets your clients book appointments, syncs to your calendar, and sends reminders.

Component Monthly cost
Server (VPS, shared across users)~$0.02/user
Database storage~$0.01/user
Email delivery (reminders, confirmations)~$0.05/user
Calendar API integrations~$0.00
TLS, DNS, CDN~$0.01/user
Total infrastructure cost~$0.09/user

Nine cents. The market rate for this tool is $12 per user per month. That's a 133x markup over infrastructure cost.

Some of that markup is legitimate. You need people to build features, fix bugs, handle support. But here's what's changed: the cost of building features and fixing bugs has also collapsed, because of the same AI tools we mentioned above. And support costs drop when your software is simple and reliable.

Where the money actually goes

If it doesn't cost $12 to serve you, what are you paying for? Mostly three things:

SaaS gross margins typically run 75-90%. The company selling you a $12/month scheduling tool is keeping $9-$11 of that after infrastructure costs. This is the gap.

What happens when you close the gap

If you price software at infrastructure cost plus a transparent margin for maintenance, something interesting happens. The same scheduling tool costs $0.75 per month. Email marketing for a small list costs $4-5. A professional desktop IDE costs $29 per year.

These aren't theoretical numbers. They're what we charge at The Commons.

This works because of a virtuous cycle:

  1. AI reduces development cost
  2. Lower development cost means lower prices
  3. AI maintains the software (monitoring, patching, optimization)
  4. Lower maintenance cost keeps prices low
  5. When infrastructure costs drop, prices drop too

It's not a race to zero. Infrastructure has real costs. Maintenance takes real work. But the honest price for most software tools is a fraction of what the market charges.

Why now

Every software engineer knows this is happening. The conversations in engineering organizations right now all circle the same realization: if AI can write most of the code, and cloud costs keep falling, and open-source keeps maturing, then the cost structure of the entire software industry is fundamentally different than it was five years ago.

The incumbents can't act on this. They have hundreds of employees, venture obligations, and pricing structures that shareholders expect to go up, not down. They will add AI features to justify keeping prices the same. They won't pass the savings through.

New infrastructure can. If you start with the assumption that one person plus AI can build and maintain what used to take a team, and you don't carry the overhead of a sales org, a marketing department, and a board expecting 30% year-over-year growth -then you can price honestly.

What this means for you

If you're a freelancer, therapist, consultant, small agency, or anyone running a business on software tools: the tools you need don't have to cost what they cost. The economics changed. Some of us are building on the new economics.

If you're a software engineer watching this unfold: you already know. The cost of building things collapsed. The question is what gets built with that fact as a starting assumption, rather than as a threat to manage.

The Commons prices every product at infrastructure cost plus a transparent margin. When our costs go down, your price goes down. That's it. That's the whole model.

See what's available today: Products · Pricing breakdown