The Signal
The real switching cost of a vendor is rarely visible when the contract is signed. It appears when the business tries to leave.
Data exports are incomplete. Automations point at old endpoints. Credentials are shared. Historical reports have no replacement. The notice window has already started. A service owner knows how the tool works but not how to shut it down. The vendor is inexpensive on the invoice and expensive in the operating system.
A vendor exit card makes that dependency visible early. It is a one-page record of the contract window, data return, integrations, identities, replacement owner, continuity plan, customer impact, shutdown authority, and evidence required to close the relationship.
Why this matters now
Teams can add a cloud tool, contractor, data provider, agency, or specialist platform faster than procurement can understand the dependency. A free trial becomes a workflow. The workflow becomes a source of record. A source of record becomes something the business cannot replace before renewal.
Consolidation, price changes, security reviews, product shutdowns, and strategy changes can all force the exit question with little warning. The company then learns that download access is not the same as a usable migration, deleting the account is not the same as revoking every token, and contract termination is not the same as confirmed data disposal.
NIST's supply-chain risk management guidance treats supplier relationships as a lifecycle risk, not a one-time buying decision. That is the useful operating frame. Exit readiness belongs beside selection, contracting, access, and ongoing review.
The mistake to avoid
The mistake is making the exit plan a contract clause nobody has tested. Language about data return and deletion matters, but it does not prove the export contains relationships, attachments, audit history, configuration, or metadata needed by the replacement system.
Another mistake is assigning the whole exit to IT. Technology may handle identity, integrations, and migration. The business owner must decide what work continues, which history matters, what customers need to hear, and what can be retired instead of rebuilt. Finance must know the notice and payment terms. Legal or security may need deletion evidence. One team cannot infer all of that during the final week.
Run the exit-card test
The first section is commercial: contract owner, renewal date, notice deadline, termination method, outstanding commitments, and final invoice rule. Put the notice date on the operating calendar, not only in the contract repository.
The second section is continuity: business process supported, customer promise affected, replacement owner, fallback procedure, acceptable downtime, and historical records required after exit. If the vendor disappeared tomorrow, this section should tell the team which work stops first.
The third section is technical: systems of record, integrations, API keys, service accounts, single sign-on groups, webhooks, domains, automations, stored files, and downstream reports. Name who can disable each connection and how the team will know it is no longer active.
The fourth section is closure: export format, sample validation, retention obligation, vendor deletion process, backup treatment, subcontractor handling, credential revocation, asset return, and evidence of completion.
The diagnostic test is practical. Export one representative record and ask a person outside the current tool to use it. If the file opens but the business meaning is gone, the export is not exit-ready. Then revoke one nonproduction credential and verify the dependent workflow fails where expected. An exit plan should survive contact with the system.
What stays protected
Protect continuity before optimizing for speed. Do not cancel access before the replacement owner confirms the required records, workflows, and customer obligations can continue. Keep a defined rollback window when the risk justifies it.
Protect data boundaries too. A migration is not permission to spread vendor data into unmanaged folders. Use an approved destination, limit temporary access, record checksums or counts when useful, and set an end date for migration copies.
Finally, protect the right to retire work. Not every dashboard, automation, or historical field deserves to survive the exit. The card should distinguish must replace, must archive, and stop. Otherwise the business pays to recreate old habits in a new platform.
The first move
Pick one vendor tied to revenue, delivery, customer data, or identity. Draft the exit card from the live system and the signed agreement. Do not rely on the account owner's memory.
Mark every blank as a risk: unknown notice window, untested export, unnamed shutdown authority, hidden integration, missing fallback, or unclear deletion evidence. Rank the blanks by how quickly they could interrupt a customer promise.
The move this week
Run a 45-minute tabletop exit with the business owner, technical owner, finance, and whoever owns data risk. Use a real departure date. Walk from notice through final deletion evidence.
Complete one operating receipt before the meeting ends: a usable sample export, a list of live credentials, a tested fallback, or a calendar reminder before the notice deadline. The goal is not to threaten the vendor. It is to keep the company capable of making a clean choice when the economics, risk, or strategy changes.