Vendor Lock-In: How to Plan Your Exit Before You Sign a Software Contract
November 11, 2026 · 5 min read · Etzal Group Team
Choosing software is an optimistic act. The demonstration went well, the price is acceptable and everyone wants to start. Nobody at that table wants to discuss leaving. Yet vendor lock-in is decided at this moment, in clauses and technical details that are far easier to negotiate before a signature than after three years of data have accumulated.
What lock-in is made of
Dependence on a vendor comes from several sources:
- Data. It sits in the vendor's format, in the vendor's storage.
- Process. Staff have learned the product's way of working, and procedures are written around it.
- Integrations. Other systems are connected to its particular interface.
- Customisation. Reports, workflows and rules were built inside it and exist nowhere else.
- Contract. Long terms, automatic renewals and exit fees.
- Knowledge. Only the vendor understands how your instance is set up.
Each can be reduced. None can be removed entirely, and that is acceptable.
Questions to ask before signing
About data
- Can we export everything, including attachments and history, at any time, without asking?
- In what format? Open formats such as CSV, JSON or a standard database dump are useful. A proprietary backup that only the vendor's product can read is not.
- Is there an API that covers all the data, and are there limits on its use?
- Where is the data stored, and in which country?
- After termination, how long is it kept and when is it deleted?
About ownership
- For custom software: who owns the source code and the designs? Is the code delivered to our repository as it is written?
- For SaaS: who owns the configurations, templates and reports we create?
- Are the accounts for domains, cloud hosting and third-party services in our name?
About the contract
- What is the minimum term and the notice period?
- Does it renew automatically, and how much warning do we get?
- How can prices change at renewal, and is there a cap?
- What help with transition is included at the end, and at what cost?
- What happens if the vendor is acquired or stops trading?
About dependence
- Could another company support or host this?
- Does it rely on proprietary components that only this vendor supplies?
- How are integrations built, and would they survive a replacement of the core system?
The exit plan
An exit plan is one or two pages, written at purchase and reviewed yearly. It contains:
- What we would need to take with us. A list of data sets, documents, configurations and integrations.
- How each would be extracted. The export function, API or request process, and who can run it.
- Where it could go. Two or three realistic alternatives, including building or hosting it ourselves.
- How long a move would take, in rough terms, and what it would cost.
- What would trigger it. Price increases beyond a limit, repeated outages, a change of ownership, loss of a required feature.
- Who decides.
Writing it takes an afternoon. Its main effect is that you ask the right questions while the vendor is still eager to answer them.
Test the export
A promise of data portability is worth what a test shows. Within the first months:
- Run a full export.
- Open the files and confirm that records, relationships and attachments are present and readable.
- Check that a technical person could load them into another system.
- Note how long it took and what was missing.
Repeat once a year. Many organisations discover at this step that the export excludes attachments, audit history or custom fields, which is far better to learn early.
Practical ways to stay portable
- Keep your own copy. A scheduled export to storage you control serves as both a backup and an exit route.
- Prefer standards. Standard file formats, standard protocols, standard authentication.
- Limit deep customisation. Each custom workflow inside a product is one more thing to rebuild elsewhere. Configure where possible and customise where it really matters.
- Put a layer between systems. When other systems connect through an integration layer of your own, replacing one product does not mean rewriting every connection.
- Document your setup. Keep a description of configurations, rules and reports outside the product.
- Own your accounts for domains, hosting and third-party services.
- Keep knowledge in-house. At least one of your own people should understand how the system is configured.
When some lock-in is worth it
A product that fits well, is reliable and is fairly priced earns the dependence you place on it. Using its advanced features can be the right decision. Make it knowingly: record which features tie you in and what replacing them would involve.
The same applies to cloud platforms. Using a provider's specialised services can cut development time considerably. Decide deliberately which parts of the system may depend on them.
Signs that the risk is rising
- Price increases at each renewal with little explanation.
- Export functions removed or moved to a higher price tier.
- Slower support and fewer updates.
- A change of ownership followed by a change of direction.
- Requests for longer commitments in exchange for prices you previously had.
Any of these is a reason to open the exit plan and update the cost of leaving.
For custom software
The same principles apply with more force. Hold the source code, the documentation, the credentials and the deployment instructions. Make sure another team could run and change the system. A vendor relationship in which you are free to leave is usually a better relationship.
We write handover and exit provisions into our own Enterprise Software projects, and we help clients assess the exit path from systems they already use. If a renewal is approaching and you are unsure how dependent you are, talk to us.
Summary
Lock-in comes from data, process, integrations, customisation, contract terms and knowledge. Before signing, confirm that you can export everything, that you own what you pay to create, and that the contract describes the ending. Write a short exit plan, test the export, keep your own copy, and review the position every year.
Frequently asked questions
Is vendor lock-in always a bad thing?
No. Some dependence is the normal price of using a good product, and avoiding it entirely means building everything yourself. The aim is to know how dependent you are, to keep your data portable, and to have a realistic path out if price, quality or the vendor's direction changes.
What is the single most important protection against lock-in?
The ability to export all your data in a usable format, proven by actually doing it. Contracts, escrow and promises help, and a complete export that you have tested is what lets you move.
What should a contract say about the end of the relationship?
It should state how long you can access the system after termination, the format and timing of the final data export, any help with transition and what it costs, and when your data is deleted. Leaving these to goodwill is a risk.
Related service
Free consultation
Not sure where to start? We will map the right system for your business, free.
