What a supplier review should change

An account manager can deliver a thorough presentation without the meeting changing anything. The service levels are shown, the roadmap is discussed and both sides agree to keep communicating.

For an important supplier, that is not enough. I would want the review to settle a service issue, test a risk assumption or inform an upcoming commercial choice. The outcome should be clear before the next invitation arrives.

The review is also an opportunity to inspect the organisation's own contribution to the service. A supplier cannot resolve an approval delay or ambiguous requirement that the customer refuses to own.

Build a shared account of the service

Bring the supplier's measures and the organisation's evidence together. Compare incident records, user impact, unresolved problems and planned changes. Agree the period covered and the meaning of the measures before arguing about performance.

A supplier may meet its response target while users still experience a long interruption. Response time and restoration are different things. Likewise, a component may be available while the complete business process is not working.

Ask for examples behind aggregate figures. Which incidents caused the most disruption? Which problem returned? What was excluded from the measure? The objective is a reliable account of the service, not catching the account manager with a misleading slide.

Include good performance where there is evidence for it. Understanding what worked can help preserve the right capability during contract changes. A review that only lists complaints can miss useful dependencies and discourage candour.

Separate a repair from an unresolved cause

A closed incident confirms that the immediate issue was handled according to the applicable process. It does not necessarily establish that the cause has been removed.

In a fictional arrangement, a supplier repeatedly restarts an integration after it stalls. Each incident is resolved within the contracted target, but staff still reconcile delayed transactions. The review should ask what would reduce the recurrence and who owns that work.

The answer may involve the supplier, internal configuration or a third system. Agree the evidence needed to distinguish those possibilities. Do not let uncertainty become a reason for everyone to leave the issue with someone else.

Set actions with a deliverable, owner and date. "Investigate stability" is weak. "Provide the failure analysis and proposed options before the next release decision" creates something the customer can assess.

Check the effect at the next review. An action marked complete should have evidence relevant to the original problem, rather than merely a statement that engineering has reviewed it.

Examine what the contract encourages

Commercial terms influence which work receives attention. If the agreement rewards quick restoration but says little about recurring problems, discuss whether the service arrangement supports the outcome the organisation needs.

Do not assume the supplier must perform every improvement within the existing fee. Read the contract and distinguish an unmet obligation from a new request. Obtain procurement or legal advice where interpretation has material consequences.

Ask whether reporting measures still describe the business service. A metric chosen when the system was smaller may no longer reflect its use. Changes should be agreed and documented, not informally substituted because the existing measure is inconvenient.

Review incentives on both sides. Internal teams may delay approvals, change scope or send work through several channels. Correct those behaviours as part of the plan. Paying a supplier more will not necessarily resolve a customer-side decision bottleneck.

Keep ownership and exit options visible

Name the internal service owner and the people who can make commercial decisions. An account relationship should not depend entirely on one friendly contact at either organisation.

Review upcoming notice dates, renewals and changes in service scope. Start significant decisions early enough to assess alternatives. A review held after the notice deadline may document dissatisfaction without preserving any practical choice.

Inspect exit arrangements before deciding whether to use them. What data, configuration and operating knowledge can be transferred? What assistance is included? What would the organisation need to provide itself? Avoid assuming that a different supplier can take over simply because both sell a similar service.

The question in deciding what should stay in-house remains relevant even with a good supplier: the organisation needs enough internal capability to judge the work and direct a transition.

End with a decision record both sides recognise

Summarise what changed. That could be an agreed corrective action, a revised responsibility, a commercial proposal to assess or a decision to continue without change because the evidence supports it.

Distinguish agreed commitments from requests awaiting approval. Send the record promptly and resolve disagreements about it while the discussion is still fresh. Otherwise the next meeting may begin with competing memories of what was promised.

Escalate repeated non-delivery through the contractual and management routes available. A review process should not keep renewing the same action indefinitely. Explain the consequences and obtain appropriate advice before taking a consequential contractual step.

Feed relevant findings into future subscription and procurement decisions. A lesson about export limitations or unclear support boundaries should influence the next purchase, not remain trapped in one supplier's minutes.

Before closing the meeting, ask each owner to confirm their next action and what evidence will show it is complete. If the review produced no decision, no changed action and no useful assurance, shorten the next one or reconsider why that meeting exists.