Post-Project Window Supplier Scorecard: Price, Delivery, Quality and Recovery

1. Approve the scoring design before opening the outcome data

Name the decision the scorecard will inform: preferred-bidder status, conditions for another order, due-diligence depth, development actions, order-size cap or no automatic consequence. Identify review owner, contributors, supplier fact-check opportunity, approval authority, retention, confidentiality and challenge route. The scorecard should inform governance, not secretly make an award decision.

Define dimensions and evidence before seeing performance. Common lanes are commercial control, delivery, quality, recovery/claims, communication/document control and compliance/safety where relevant. The buyer sets weights, critical overrides and thresholds according to its risk appetite. A weighted total must not allow an unsupported safety, fraud or certification concern to be neutralised by low price unless approved governance expressly permits that treatment.

Use score descriptors anchored in observable states rather than adjectives. For every band, state required evidence and boundary. Record ‘not applicable’, ‘unknown’ and ‘insufficient evidence’ separately from zero performance. Do not reward missing records by excluding difficult metrics after the project.

2. Rebuild baseline budget, approved changes and final cost

Start with the accepted quotation/contract and line-item scope—not the first enquiry or final invoice. Reconcile opening marks, quantities, systems, glass, hardware, finish, screens/accessories, samples/tests, packing, freight, services, exclusions, Incoterm/named place, currency and taxes. Record provisional or buyer-supplied items separately.

Create a change ledger linking each authorised variation to trigger/date, affected marks and baseline revision, initiator, reason, added/omitted work, agreed price/time treatment, evidence and status. Separate buyer/design changes, supplier proposals, unforeseen verified conditions, error correction, acceleration and disputed items without converting category into legal fault.

Compare final authorised cost with the adjusted baseline, then isolate unapproved, disputed, avoidable rework, duplicate freight, credits and unresolved claims. Currency and index movements need their contract-defined basis. A low final total can reflect omitted scope; a high total can reflect buyer additions. Score commercial control from transparent causation and forecast discipline, not raw variance alone.

3. Compare promised and actual delivery through prerequisite-aware milestones

Freeze the authorised programme baseline and revisions. For each group record prerequisite dates—deposit, drawings, samples/colour, glass/hardware decisions, compliance evidence and site inputs—then promised material, first-off, production, inspection, packing, booking, shipment and delivery events. Distinguish forecast, commitment and actual evidence.

Measure by opening group or shipment, not one project completion date. Record missed events, duration and downstream consequence only from evidence. Separate supplier-controlled production/logistics, buyer/design input delay, carrier/customs, force-majeure or unresolved causes according to approved project analysis; this is not a contractual entitlement decision.

Evaluate forecast quality and communication: Was the plan logic visible? Were risks disclosed before commitment failed? Did revised dates include cause, affected marks and recovery plan? A supplier can be late yet manage recovery transparently; another can meet a final date through unsafe compression or repeated unreported slippage. Preserve both outcome and control behaviour.

4. Score quality and recovery from populations, recurrence and verified closure

Build a quality denominator: total units/parts, inspected population and method, characteristics accessible, shipment/site populations and uninspected scope. Reconcile NCRs, shortages, damage, wrong configuration, documentation and service cases by unique ID and affected opening marks. Raw issue count without denominator or duplicate control is misleading.

Classify only under the approved quality plan. Record requirement source, observation, affected population, containment, disposition, reinspection and closure. Separate isolated damage from common process or repeat failure. A repaired unit can close locally while recurrence control remains open; a large shipment with few inspected units should not receive a confident low-defect claim.

Evaluate recovery as a distinct dimension: acknowledgement, containment, evidence preservation, root-cause quality where required, technically authorised remedy, replacement/configuration accuracy, freight/site coordination, communication, verified closure and claim/credit resolution status. Fast unsupported action can worsen evidence or repeat the wrong part. Score response from approved expectations and actual records, not a universal clock.

5. Apply confidence, review contradictions and convert scores into bounded actions

Assign evidence confidence separately from performance: high when primary records reconcile and coverage matches the claim; medium when evidence is partial but corroborated; low when based on recollection, selected messages or unclear populations; unknown when missing. The buyer approves definitions. Never multiply numbers mechanically if the method obscures a critical missing denominator.

Triangulate buyer, supplier, inspector, carrier and site records. Invite correction of factual identities, dates and documents, preserving the audit trail. Supplier disagreement belongs in the final record with evidence and disposition; it does not automatically raise or lower the result. Separate disputed facts from different commercial interpretations.

Issue a signed outcome containing dimension results, weights, confidence, critical overrides, strengths, weaknesses, open matters and authorised next-order controls: smaller first order, extra evidence gates, revised packaging, named project lead, corrective plan, credit review or other buyer decision. Do not publish allegations or blacklist without legal/procurement authority. Revalidate current supplier facts before a future award.

FAQ

Continue the project review

Connect this guide with the proposed window and door systems, then review how Cavorlen approaches manufacturing and quality control. When your schedule is ready, send it through the project review form.

Related buyer guides

• A Contract Line-Item Schedule for Window and Door Packages
• Window Non-Conformance Reports: Containment, Rework and Reinspection
• A Window Factory Quality-Control Checklist Buyers Can Actually Use