You do not need to threaten termination every time a testing supplier disappoints you. Replacing a supplier is expensive, disruptive and sometimes just swaps one set of problems for another.
Start by making the problem specific enough that both sides can do something about it. Challenging a supplier does not require beginning the meeting with ‘we are considering termination’, unless your procurement team enjoys cardio.
If the relationship has become too political for a sensible conversation, we can provide an independent supplier review without beginning from the assumption that the answer is replacement.
Challenge evidence, not personalities
Instead of ‘the testing is poor’, show what is happening. Serious defects repeatedly escape. Regression takes too long. Reports hide blocked tests. Automation consumes effort but catches little. Senior supplier staff are absent. The team cannot explain risk.
Specific evidence creates a fixable conversation and makes contractual escalation much easier if improvement does not follow. Escalation becomes more credible once you can answer whether to replace the testing supplier or fix the existing relationship.
Agree the outcome you need
Do not respond to a poor service by adding more supplier metrics unless those metrics help a decision. Agree what must improve: faster feedback, stronger coverage of critical journeys, better defect quality, clearer release reporting, more stable automation or better senior challenge.
Then measure that outcome directly where possible.
Fix incentives that reward the wrong thing
If the supplier is commercially rewarded for volume, do not be surprised when it optimises volume. Contracts built around test counts, utilisation or automation percentages can produce exactly the behaviour management later complains about.
Change the conversation from ‘did you meet the metric?’ to ‘did the service help us make safer and faster decisions?’
The goal is to make the service better, not to win an argument and inherit three months of passive-aggressive status reports.
Make improvement time-bound
Agree a small number of changes, owners and dates. Review whether the actual service improved. Do not let the recovery plan become another permanent workstream whose status is always amber.
If the supplier responds well, you may preserve valuable product knowledge and avoid a painful transition. If it cannot improve despite clear evidence and a fair opportunity, replacement becomes a much easier decision to justify.
Use the contract as a tool, not the whole conversation
Contractual obligations matter, especially where performance has genuinely fallen below what was bought. But hiding behind clauses on either side can stop the more useful question: what service would actually work better?
You may need to change roles, governance, measures, incentives or the split between internal and supplier responsibility. Some of those changes can be made without a full procurement exercise.
Challenge should therefore have two tracks: hold the supplier to commitments that still matter, and jointly remove commitments that now drive pointless behaviour. A bad metric remains bad even when somebody is contractually obliged to hit it.
Make senior supplier leadership own the recovery
If the issue is serious, do not leave recovery entirely with the same delivery layer already struggling. Ask senior supplier leadership to explain the cause, proposed changes and how they will know the service has improved. If evidence fixes the relationship, good. If evidence makes the relationship worse, that is also useful evidence.
That creates accountability above individual testers and gives the supplier room to change staffing, incentives or delivery structure where necessary. It also stops every problem being pushed down into ‘team coaching’ when the model itself may be wrong.