Review NDT Supplier Performance with Useful Evidence

Define the purpose of the supplier review

An NDT supplier review should support a concrete decision: renewing an engagement, improving a recurring workflow, resolving a service problem or deciding what evidence to request before future work. State that purpose before collecting scores. Otherwise the review can become a collection of convenient numbers that do not help the buyer or provider improve anything.

Define the period, projects and service scopes included. A provider performing one unusual investigation should not be compared casually with a provider completing many repetitive production batches. Record the context so that later readers understand what the review does and does not describe. A favorable conclusion for one scope is not universal approval for every method, location or customer requirement.

Separate service quality from component condition

An inspection provider may report findings that create additional work for the buyer. That outcome does not by itself indicate poor service. Conversely, a report with no relevant findings is not proof of excellent provider performance. The condition of the examined items and the quality of the inspection process are different matters.

Avoid measures that reward a supplier for producing fewer findings, less escalation or fewer questions regardless of context. Such incentives can distort the purpose of the service. Review whether the supplier followed the agreed scope, communicated limitations, produced usable evidence and supported the required technical process. Questions about interpretation or examination validity need qualified assessment, not a purchasing score chosen for convenience.

Begin with the commitments that existed

Collect the agreed scope, quotation clarifications, delivery milestones and approved changes for the reviewed jobs. A performance judgment should relate to an obligation that the parties can identify. If a buyer expected a particular data format but never included it in the engagement, treat that as a requirement-definition gap rather than automatically labeling the supplier's delivery defective.

The IAQG's public supplier-management overview connects selection, requirements and performance monitoring within its aerospace supply-chain context. [1] The useful application here is to keep the review connected to the actual purchasing and quality requirements. This guide does not impose aerospace requirements on unrelated work or establish compliance with any particular management-system standard.

Choose a small set of decision-relevant measures

Possible review areas include agreed delivery timing, report completeness, responsiveness to technical questions, accuracy of identifiers and closure of verified service issues. Select only measures that the organization can define and support with evidence. A short set of reliable measures is more useful than a large dashboard whose entries require subjective reconstruction.

FEMP's maintenance guidance discusses selecting indicators in relation to operational and organizational objectives. [2] Apply that discipline by asking what action each proposed supplier measure would change. If a number cannot influence a decision or prompt a useful investigation, consider whether collecting it is worth the effort. Do not add it merely because the available software can display a chart.

Define the denominator and event

A percentage needs an explicit denominator. For report timeliness, decide whether the unit is an agreed report package, an individual report or a complete purchase order. Define the deadline event and the delivery event. Sending an incomplete file may not meet the agreed deliverable, while a technically complete package awaiting buyer administration may require a different interpretation.

Write these definitions before calculating results. Retain the underlying job list so that someone can reproduce the number. If the definition changes, show the change and avoid presenting the new result as directly comparable with the old series. Apparent improvement can arise from changing what is counted rather than improving the service itself.

Preserve both timing and cause

Record the actual dates and the agreed dates without altering them to make the story simpler. Then record the cause of significant differences separately. A report may be late against the original plan because the buyer added scope, because a required input was missing or because the provider failed to meet its commitment. Those are materially different situations.

Allow an unresolved attribution when the evidence is incomplete. Forcing every variance into buyer fault or supplier fault can create false certainty. Assign an owner to investigate and retain the basis of the eventual decision. The review should make accountability clearer without rewriting the event history to fit a preferred conclusion.

Examine report usability beyond appearance

A polished report can still be difficult to use if it has ambiguous item identity, missing referenced attachments or unclear status. Review a sample against the agreed delivery requirements. Ask whether the receiving organization could link the report to the correct job, understand its limitations and locate the supporting evidence needed for its decision.

Separate administrative defects from technical concerns. An incorrect purchase-order reference may be corrected through a defined administrative route. A question affecting examination interpretation or acceptance needs the appropriate technical review. Combining every issue into a single defect count can conceal their different consequences and encourage attention to easy corrections while more important questions remain unresolved.

Use sampling with visible limitations

A buyer may not be able to inspect every aspect of every completed engagement during a periodic review. If sampling is used, record how the sample was chosen and what it covers. Include enough context to avoid implying that a small convenience sample demonstrates the condition of all work.

Do not select only the best-known success or the most frustrating failure. A practical review may examine different service types, delivery periods and issue histories. The appropriate sampling approach depends on the organization's quality requirements and the purpose of the review. This guide does not supply an audit sampling rule or a statistical confidence claim.

Worked example: a misleading late-report score

In a fictional review, a buyer finds that four of ten report packages arrived after their original planned dates. A simple dashboard labels the supplier's on-time result as six out of ten. The underlying records show that two packages had approved deadline changes after buyer-added scope, one awaited a missing drawing revision and one missed an unchanged supplier commitment.

The review preserves the original date comparison but adds the agreed-change history and cause analysis. The buyer can now address its own scope control, resolve the drawing handoff and discuss the supplier's missed commitment separately. The example uses invented counts to explain the method; it is not a benchmark, a real supplier result or a recommended scoring threshold.

Invite the supplier's evidence before deciding

Share the factual review package with the supplier and allow a defined opportunity to correct errors or provide missing context. Ask for specific records rather than broad disagreement. The supplier may identify an approved change, a delivery acknowledgement or a technical query that the buyer's dataset omitted.

Keep the review owner responsible for evaluating that evidence. Supplier participation does not mean every concern must disappear by consensus. Where the parties disagree, record the positions and the decision route. A fair process strengthens the resulting action because both sides can see how the conclusion was reached and which evidence remains contested.

Turn verified problems into precise actions

An action such as improve communication is difficult to implement or verify. Identify the failure mechanism and the intended change. For example, if report questions repeatedly reach an unattended mailbox, the action could establish a named review contact and an acknowledged transfer when that person is unavailable.

Assign the action to someone with authority to make the change and agree a completion date appropriate to the issue. Distinguish immediate correction of the affected job from improvement of the underlying process. Correcting a filename does not necessarily prevent future misidentification; the broader action may need to address how identifiers are generated and checked.

Check effectiveness on later work

An action is not proven effective merely because a new form or instruction exists. Decide how the next relevant engagement will demonstrate whether the change works. The check might examine a sample of handoffs, the handling of a report query or the completeness of a revised delivery package.

Record the result and any remaining weakness. If no suitable job occurs during the planned period, say that effectiveness is not yet demonstrated rather than closing the action automatically. This prevents an improvement register from becoming a list of completed documents with no evidence that day-to-day service has changed.

Avoid scoring away a material concern

A weighted score can hide an important unresolved issue behind strong performance in unrelated categories. A supplier might communicate promptly and invoice accurately while a significant technical concern remains open. Decide which matters require separate disposition and cannot be offset by points elsewhere.

The appropriate decision belongs to the organization's authorized quality, technical and purchasing roles. This guide does not define automatic suspension thresholds. It recommends preserving the concern, its evidence and the decision separately from an aggregate rating. Anyone approving the next engagement should be able to see the unresolved issue without decoding the arithmetic behind a dashboard.

Compare like work and explain differences

When comparing providers, consider differences in scope complexity, customer readiness, location, report requirements and volume. A supplier handling difficult one-off assignments may generate more clarification traffic than a supplier processing stable batches. More questions can reflect a different task rather than weaker capability.

Use comparison to identify useful questions, not to declare a universal ranking from thin evidence. Explain which differences have been adjusted for and which remain. If the dataset is too small or inconsistent for a meaningful comparison, present a narrative review of each engagement instead of a precise-looking league table. The reader needs an honest basis for judgment.

Keep commercial disagreements distinct

An invoice dispute may involve scope interpretation, authorization or purchasing administration. It should be investigated using the relevant commercial records. Do not let it silently alter the assessment of technical examination quality. Equally, a favorable price does not resolve a technical concern or compensate for an unusable report.

Where a disagreement spans both areas, involve the appropriate roles together. For example, additional attendance may follow a technical scope change that was never reflected in the purchase record. The useful action may require better coordination between the buyer's technical and commercial teams rather than a penalty applied to the supplier's overall score.

Use the worksheet to build an evidence-led meeting

Complete the review boundary and agreed commitment rows before calculating any measure. Add references to the source records, then describe the observed variance and its context. Leave attribution open when it is not yet established. This makes the meeting an evidence review rather than a negotiation over unexplained grades.

During the discussion, capture the supplier's response and the accepted next action. Afterward, send the controlled record through the organization's normal process so that participants can identify factual corrections. The worksheet is a working aid, not a certification, audit report or universal supplier approval form. Adapt it to the actual engagement and quality arrangements.

Record the decision for the next engagement

The review should end with a usable decision: continue under existing arrangements, request additional evidence, implement an improvement action or refer an unresolved matter to the authorized approval process. State the scope of that decision and any conditions. Do not let a general green status imply approval beyond the work examined.

Before a new job, still confirm the actual requirement, available resources and relevant supplier approvals. NDT Connect can assist discovery and service enquiries, but a listing or prior positive engagement does not replace this project-specific assessment. Retain the review as one piece of evidence in a continuing relationship rather than treating it as a permanent guarantee of future performance.

Control access to candid review records

Supplier review notes can contain commercial information, personnel observations and unresolved allegations. Share them with the people who need them for the defined business purpose, using the organization's normal access and retention arrangements. A private improvement discussion should not become an unsupported public claim about a provider through an exported spreadsheet or a casually forwarded email.

Write observations in factual terms. Identify the affected engagement, the evidence and the current status rather than assigning motives to individuals. If a concern is later corrected or disproved, update the controlled review record and notify the relevant decision makers. Preserve enough history to explain the change without leaving a superseded accusation as the apparent current conclusion.

Keep individual personnel qualification decisions in their appropriate technical and employment processes. A supplier scorecard is not a substitute for those processes, and broad access to personal documents is not necessary simply because a periodic business review is taking place.

Make the buyer's improvement visible too

Supplier reviews are more credible when the buyer records its own contribution to recurring problems. Late drawings, unclear data requirements and inaccessible decision makers can undermine a capable provider's work. Assign buyer actions with the same clarity used for supplier actions and follow them through the same review cycle.

This does not reduce the supplier's responsibility for its own commitments. It makes the entire service arrangement easier to improve. A review that only allocates blame may produce defensive reporting, while a review that identifies controllable causes gives both parties a practical way to reduce avoidable delays and deliver clearer examination evidence on the next job.