Open source versus commercial SaaS

Open source drops the licence and adds the operations: hosting, patching, backups, monitoring and a named person who answers when something breaks. Commercial SaaS drops the operations and adds a fee that grows with seats, contacts, events or tasks. Which is cheaper depends on how fast you hire, how specific your process is, and whether anyone will own a server.

This is the comparison we hand a client before quoting.

Which five lines decide it?

Five lines: the licence against the operations, growth, fit to your process, where the data sits, and what leaving would cost. Each leans differently, so the answer changes from tool to tool.

• The subscription is a known monthly number. Ownership is a smaller monthly number plus attention.
• A per-seat fee compounds with every hire. Hosting does not.
• The more your process differs from what the product assumes, the more the subscription costs you in workarounds.
• Self-hosted keeps the data where you put it. Hosted puts it under someone else's terms.
• Export formats, missing joins and rebuilt integrations set the real switching price, on either side.

What does the open-source side actually cost?

A server sized to the load you have, one deployment done properly, security patches and version updates, backups with a restore you have tested, monitoring that alerts a human, and somebody on the team who knows how the pieces fit. Leave out the last one and the rest becomes a liability.

We would rather say that plainly than sell ownership as a way to run nothing. Open source versus SaaS total cost of ownership works through them one by one.

Where does commercial SaaS win outright?

Where no mature open-source equivalent exists, where the work is regulated and a certified vendor is required, or where the team has nobody to operate a server. In those cases paying the vendor costs less, and telling a client otherwise wastes their money.

Where does open source win outright?

On load-bearing systems that have a mature open-source equivalent, where the per-seat bill compounds as you hire, where the process is specific enough that the product's limits are the constraint, or where the data cannot leave your accounts.

That is the test we use when deciding which systems are worth migrating.

How do you decide in an afternoon?

List every tool, its annual cost and who depends on it, then mark the load-bearing ones. For those, check whether a mature open-source equivalent exists. Price a migration only for the ones where it does, and set that against the subscription over the period you intend to keep it.

Then keep the rest rented. Most businesses end up running a mix, which keeps the migration budget on the systems that carry the load. Open-source business systems covers selection, deployment and handover.