Vendor lock-in is not simply using a product for a long time. The warning appears when leaving would be unexpectedly difficult, slow, expensive, or damaging to records and operations. For project management software, test data portability, workflow reconstruction, integrations, identity, contract terms, and staff knowledge before the system becomes critical.
Treat export as a test, not a checkbox
Request a representative export before purchase or renewal. Include projects, tasks, custom fields, comments, attachments, status history, dependencies, users, permissions, boards, reports, automation, forms, and archived work. Open the files with ordinary tools and map each important field to a plausible destination.
Atlassian's Jira Cloud export documentation shows why the detail matters: its site backup can include work items, custom fields, boards, sprints, users, groups, comments, media, and configuration, while some features such as automation flows and third-party app data are not included automatically. Every candidate product has different boundaries, so verify the current export rather than assuming the word `backup` means complete portability.
Red flags include an export available only after cancellation, undocumented formats, missing identifiers or relationships, attachments represented only by temporary links, no incremental export, administrator-only access with no backup owner, and a file that the provider cannot demonstrate importing anywhere.
Inventory proprietary dependencies
List custom workflows, formulas, automation rules, dashboards, marketplace apps, identity mappings, notification channels, and API integrations. Rate each as portable, reconstructable, replaceable, or unknown. A standard task table may move easily while years of workflow logic do not.
UK government technical lock-in guidance recommends monitoring switching costs and, for SaaS, choosing open standards and formats so data can be exported to a new environment. It also recognizes that some lock-in can provide value. The decision should therefore compare the product benefit with tested exit cost instead of demanding zero dependency.
Examine commercial and operational signs
Review notice periods, minimum commitments, data-egress or export charges, retention after cancellation, support required for export, API limits, and how quickly access ends. Confirm who owns custom content and whether the contract promises data return. Do not rely on a sales statement that is absent from the agreement or current documentation.
Operational warning signs include one administrator, undocumented configuration, integrations owned by personal accounts, no record of API credentials, and no tested replacement for critical notifications. Staff knowledge can be a form of lock-in when only a vendor consultant understands the system.
Run a bounded exit exercise
Choose a representative project and export it. Import or reconstruct it in a neutral staging system. Compare counts and fields, open attachments, follow dependencies, verify dates and users, and calculate the time and manual correction required. Preserve the source system; an exit test should not delete or alter production data.
Repeat the exercise after major customization or before a long renewal. Record unsupported fields and decide whether to accept, mitigate, or avoid them. A weakness may be acceptable when the product creates substantial value and the organization has a funded exit plan.
Make the decision explicit
A selection scorecard should include export completeness, tested import result, missing configuration, integration count, identity dependencies, contractual exit terms, estimated migration effort, required skills, and business value. Reject optimistic labels such as `fully portable` unless the evidence covers the workflow that matters.
The strongest warning sign is not a proprietary feature by itself. It is a critical dependency whose exit path, cost, owner, or evidence remains unknown.
Sources
- Export data from Jira Cloud — Atlassian Support
- Managing technical lock-in in the cloud — Government Digital Service and Central Digital and Data Office