Lock-in isn't a side effect of SaaS: it's the product
When you evaluate a SaaS tool, you look at the features, the entry price and the interface. The vendor, meanwhile, looks at something else: how much it will cost you to leave. That number — your exit cost — is the most valuable asset they have. Lock-in isn't a side effect of the model. It is the model.
The four classic tactics
1. The incomplete export. The button exists, so technically "you can take your data". What you take is the contacts. What stays: the activity history, the notes, the file attachments, the automations you built over years. You get the bricks; they keep the house.
2. Proprietary formats. Your data comes out in a format only that tool understands — or in a CSV so degraded that rebuilding the relationships between tables takes weeks. Technically exportable, practically useless.
3. Integrations that break. While you're in, everything connects to everything. The day you announce you're leaving, you discover those integrations were the glue of your operation: the site form, the billing sync, the webhook that triggers your pipeline. Leaving doesn't cost the subscription: it costs rebuilding half the company.
4. Prices that scale with your success. Per-contact, per-seat, per-record pricing has an elegant, perverse property: your bill grows when your business grows. You doubled your customer base, you doubled your invoice. Your growth is their revenue model. You're not paying for more software — the software is the same — you're paying for having done well.
How to spot it before you sign
You don't need faith, you need ten minutes of due diligence:
- Export on day one. Create test data and try the export before migrating anything. If what comes out isn't enough to rebuild your operation, you already know what that button is worth.
- Look at the pricing curve at 10x. Work out what you'd pay with ten times your current users, contacts or records. If the number bothers you today, it will hurt in two years.
- Read the API limits. Low rate limits and a toy API are a signal: they don't want your data leaving through the service door.
- Ask about the format. Do they document their data model? Do they export in open formats (CSV, JSON, ICS, vCard) or only in their own?
The alternative: ownership
The only structural defence against lock-in is ownership. Software whose code you can download, whose database lives on your server — or on a cloud billed at cost that you can leave in an afternoon.
That's how we run Zerosoftware: all the code is open source; if you use our cloud, you pay what the infrastructure costs, with no margin; and your data exports complete, in open formats, with one click. If we ever do things wrong, leaving must be easy. Staying should be a decision, not a sentence.