Skip to content
← Insights

Rescue or Exit: What to Do With a Failing Vendor Relationship

Most failing vendor relationships are neither rescued nor ended. They are tolerated. The client escalates, the vendor apologizes, a credit shows up on the next invoice, and twelve months later the same people are having the same meeting.

Tolerating is the most expensive of the three options, because you pay full price for partial delivery and carry the internal cost of managing around it. A real decision is cheaper in either direction. The trouble is that the decision usually gets made on mood. A bad quarter produces "we need to get rid of them." A look at the switching cost produces "we're stuck." Neither is analysis.

Here is how to make the call on evidence.

Takeaway 1: Find out whose failure it is before you decide anything

Take the last six months of misses and sort each one into three piles: caused by the vendor, caused by you, caused by the contract.

Vendor-caused misses are what most people expect to find: thin staffing, turnover, weak supervision, missed commitments. Client-caused misses are the ones nobody volunteers: late approvals, access requests that sit for a week, requirements that changed after work started, three different people giving the vendor direction. Contract-caused misses come from an agreement that pays for the wrong thing, such as tickets closed when you needed problems solved.

Example: A support vendor keeps missing its response-time SLA and the client is ready to go to market. A review of the tickets shows that a large share of the late ones were waiting on the client's own security team to grant system access. A new vendor would inherit the same queue and miss the same SLA. The fix was an internal access process, and it took three weeks.

If most of the pile belongs to you or to the contract, exiting will move the problem to a new logo and add a transition bill.

Takeaway 2: Test the vendor's capability and will with a short, hard corrective plan

Once you know which failures belong to the vendor, give them a fair and specific chance to fix those. Sixty to ninety days. Named people on both sides. Three or four measurable targets. A weekly review. A stated consequence if the targets are missed.

How the vendor responds tells you most of what you need to know. A vendor worth keeping puts senior people on the account, shares its internal data without being asked twice, and tells you which of your own habits are making things worse. A vendor to leave sends the account manager with a slide deck, disputes the numbers and asks for a contract extension in exchange for the improvement.

Example: Two vendors receive the same kind of corrective notice. The first replaces the delivery lead within two weeks, opens its quality audit results and proposes a joint weekly review. The second responds with a letter from its legal team about the client's obligations. The first relationship is recoverable. The second ended the day that letter was sent, and the only open question is the timeline.

Some findings skip the corrective plan altogether. If the vendor has misreported performance data, lost the specific capability you hired it for, or is plainly walking away from your line of business, start the exit.

Takeaway 3: Price the exit honestly, and stay ready to use it

Leaving a vendor costs more than the termination fee. Count the transition project, the months of paying two vendors at once, the knowledge that walks out with the old team, the performance dip while the new team learns your environment, and the time your own people will spend on all of it instead of their jobs.

Then count what staying costs: the rework, the customer impact, the management hours, the credibility you spend each time you explain another miss to your leadership. Put both numbers on one page. The comparison is often closer than either camp inside your organization expected.

Example: A company decides to exit and discovers during transition planning that there is no usable documentation. The procedures, the workarounds and the system quirks all live in the heads of the outgoing vendor's staff, who have little reason to write them down now. The transition runs months longer than planned, and the client pays the outgoing vendor a premium for a knowledge transfer that should have been a contract deliverable from day one.

That is the argument for exit readiness even when you choose to rescue. Own the documentation. Make sure the contract has termination assistance terms with defined pricing. Keep enough internal knowledge to brief a replacement. A vendor who knows you can leave behaves differently from one who knows you can't, and you negotiate differently too.

Making the call

Rescue when the failures are shared, the vendor responds to a corrective plan with real resources, and the cost of leaving outweighs the cost of fixing. Exit when trust in the data is gone, the capability is gone, or a fair corrective plan has already failed once.

Whatever way it goes, put a date on the decision. A relationship that has been "under review" for a year has been decided by default, and the default is the expensive option.