Skip to content
BDOT SOFTWAREBDOT Software

Insights / Product Engineering

Vendor lock-in in business software: the checks to run before you buy

Before you sign for business software, check what leaving costs. Data export, API access, contract terms and exit clauses decide whether you own your system or rent it.

Kiran Bandarupalli · 8 Nov 2026 · 5 min read

Illustration of a business owner with a key and checklist standing before a locked gate guarding a server room, with an open path to a modern office building

Every business software purchase answers two questions: does it solve today's problem, and what does it cost to leave? The first gets a demo, a trial, and a spreadsheet. The second gets skipped — yet it decides whether you own your system or rent your own data back from it. Vendor lock-in is rarely a dramatic trap. It is a series of small decisions — a proprietary export format here, a per-seat contract there — that add up until switching costs more than staying, even when the software no longer fits.

Price the exit before you sign, while you still have leverage. These are the checks that do it.

Lock-in is a price, not a feeling

Lock-in is best understood as a number: what it would cost, in money, time, and lost data, to move to a replacement next year. If it is small, you can walk away when the vendor raises prices or drops features. If it is large, you stay and pay. So the question to ask in a sales call is never “is this software good?” but “what exactly happens on the day I leave?”

Most buyers never ask because the purchase is urgent and the exit is a problem for a future quarter. But the same logic applies to the build vs buy decision: the total cost of a system is its whole lifecycle, and the exit is part of that lifecycle. A cheap subscription with a prohibitive exit can be the most expensive option on the table.

The four places lock-in hides

Lock-in rarely lives in one place. It is spread across data, integrations, contracts, and habits.

1. Your data

If you cannot get your data out in a usable form, you do not own it. Ask for a full export in an open format — CSV, JSON, or a documented schema — and check three things: does it include everything (history, attachments, and audit trails, not just current records), does it preserve the relationships between records, and can you re-import it somewhere else without the vendor's help? Treat backups you control as part of the same question: if the vendor holds the only copy of your data in a form only they can read, the exit price is whatever they decide it is.

2. Your integrations

Every integration you build on a vendor's API is an investment in staying. That is fine when the API is stable, documented, and complete. It is a trap when the API covers only the easy half of the product, or when it changes without versioning and your integrations break on their schedule. Before you commit, check whether the API can do everything your workflows need, including bulk export, and whether there is a public changelog and a deprecation policy.

3. Your contract

Read the exit clauses the way you would read the price. Look for auto-renewal terms, minimum terms that outlast your planning horizon, termination fees, and data-return deadlines — some contracts give you thirty days to retrieve your data after cancellation and then delete it. Check who owns the configuration and reports your team built inside the product, and the price-increase clause: a contract that allows unlimited annual increases lets the vendor test how locked in you are. None of this alone is a reason to walk away — but it is a reason to negotiate before signing, when the vendor still wants the deal.

4. Your team's habits

The softest form of lock-in is also the hardest to see. A tool your team has used for three years has trained them in its workflows, shortcuts, and quirks. Switching means retraining and slower work for a quarter. That cost is real and should be counted in the exit price — but it shrinks with documentation and with boundaries you can operate: the more your own processes live in your own runbooks rather than in one vendor's interface, the cheaper the move.

Run the exit test before you sign

Do not take the vendor's word for any of this. During the trial, run a small exit test: export a slice of real data, import it into a spreadsheet or a test database, and check what survives. Check these four things:

# Export a real slice during the trial, not a demo dataset, then verify:
#   1. Count records: exported rows vs. records shown in the UI
#   2. Check relationships: do orders still link to customers?
#   3. Check attachments: are files included, and do they open?
#   4. Check history: are status changes and audit trails present?
# If any answer is no, the exit price just went up — negotiate or walk away.

Run the same test for the API: pull a day of records through it and confirm the fields you need are there. If the export only works as a manual click, say so in your evaluation — for large datasets that rules out a scheduled, automated export. The same thinking applies to queued export jobs: an exit you can automate is an exit you will actually be able to use under pressure.

The pre-purchase checklist

  • Full data export in an open format, including history and attachments, tested during the trial.
  • API access that covers everything your workflows need, with versioning and a deprecation policy.
  • Contract terms you have actually read: renewal, termination fees, data-return window, price-increase limits, and who owns your configuration.
  • A named exit owner on your side: one person who knows how the export works and has run it at least once.
  • A migration budget in the business case, even if only a rough range — if the exit has no budget, it has no plan.
  • A yearly review of the lock-in picture, while switching is still a choice rather than an emergency.

When lock-in is the right trade

Not all lock-in is bad, and refusing every dependency is not a strategy. A specialist tool that does one thing brilliantly may be worth a high exit price if the alternative is building that capability yourself. The question is never “is there lock-in?” — there almost always is — but “is the lock-in priced into the decision, and is it the cheapest way to get this capability?” A niche scheduling tool with a painful export is a different commitment from your core CRM or accounting system. Match the scrutiny to the stakes, and be honest about which systems you could replace in a month and which would take a year. That honesty is also what makes the rebuild or refactor decision tractable later.

Lock-in is not something that happens to you. It is something you agree to, one unchecked box at a time. Run the exit test, read the exit clauses, and keep the price of leaving visible next to the price of staying. Then buy with your eyes open — and if the system you need does not exist yet, that is the case for custom software built around your data, where the exit is yours by design. If you want help weighing a purchase against a build, talk to us — we would rather help you avoid a bad lock-in than rescue you from one later.