Look at what a business pays for software and you will usually find the same pattern: a handful of tools doing the actual work, and a long tail of subscriptions that made sense once.
Why the price does not match the value
Most SaaS is priced per seat, per month, at a rate set by the largest customer the vendor hopes to win. That rate has to cover a feature set built for enterprises with requirements you do not have.
So you pay for a permissions model you never configure, integrations you never connect, and reporting nobody opens — because the vendor cannot afford to sell you only the part you use.
The three costs nobody counts
- Seats for people who barely log in. Most per-seat tools are paid for the whole team and used by three people.
- The annual increase. Fifteen percent compounding is invisible monthly and substantial over three years.
- The switching cost you have already accepted. Once your data is inside, the price can rise faster than your willingness to migrate.
Where SaaS is genuinely worth it
This is not an argument against buying software. Anything that is a core competency for the vendor and a distraction for you is worth paying for — payroll, accounting, email delivery. You will not build those better.
The waste is elsewhere: paying enterprise rates for a fraction of a product, and paying separately for a dozen things that could arrive as one.
What to do about it
- Audit by usage, not by opinion. Pull the last ninety days of logins for every tool. The results are usually uncomfortable.
- Count the tail. Ten subscriptions at $20 attract less scrutiny than one at $200, and cost the same.
- Look for overlap. Most stacks contain three tools that each do a bit of the same job.
- Prefer bundles where the parts genuinely belong together. Infrastructure is the clearest case — servers, domains, certificates, backups and monitoring are one job split across nine invoices.
The goal is not to spend less on software. It is to stop paying nine times for one outcome.