Shipsy turns carrier performance into a live signal that drives allocation, not a quarterly PowerPoint. Every shipment’s execution feeds a multi-dimensional scorecard; Astra, Shipsy’s planning agent, consumes those scores in real time when deciding which carrier gets the next booking. The result: performance isn’t just measured, it’s enforced automatically through allocation share.
Why we built this
Shippers managing multi-carrier networks — CEP aggregators, big retail, freight forwarders — typically run one of two broken patterns: a rigid contractual allocation split (e.g., 40/30/30) that ignores live performance, or a manual “whoever shouted loudest in the last QBR” approach. Both leak money. A leading Western European parcel operator with 50%+ national market share and a MENA retail conglomerate with 80+ years operating multi-brand portfolios both flagged allocation discipline as the biggest undermanaged lever in their logistics P&L.
Shipsy built the scorecard-to-allocation loop so performance gaps close in days, not quarters. Carriers who perform win more share; carriers who drift lose share until they course-correct.
How it works
The scorecard schema. Every carrier is scored on six dimensions, each weighted by shipper preference: - On-time pickup (pickup before cutoff) - On-time delivery (delivered within SLA) - First-attempt delivery rate (FADR) - Exception rate (damage, loss, mis-sort) - Billing accuracy (invoice matches booking) - Digital compliance (milestone updates, ePOD capture, API uptime)
Each dimension rolls up from shipment-level events captured across Shipsy’s TMS, last-mile, and multi-carrier modules. Scoring runs continuously — there’s no monthly batch.
Source-of-truth event stream. The scorecard ingests ePOD events, carrier milestone webhooks, Shipsy’s driver-app telemetry where applicable, and billing events reconciled by Nexa. When a carrier reports “delivered” but the ePOD is missing required artifacts (geofence hit, OTP, photo), it counts against the digital-compliance dimension regardless of the carrier’s own report. This prevents carriers gaming their own SLA reporting.
Weighted aggregation. Each shipper configures a weighting profile — a fashion retailer might weight FADR at 40% and billing accuracy at 10%, while a pharma CDMO weights digital compliance at 30% because of GDP audit requirements. Scores are normalized to a 0-100 scale per dimension and combined into a carrier-level composite. A global pharma CDMO handling multi-country clinical supply uses the composite to drive fully automated allocation decisions for non-critical shipments.
Allocation feedback loop. When a new shipment enters the booking queue, Astra reads: service-level requirement, lane, weight/volume, SLA tier, and carrier scorecard composites filtered to carriers available on that lane. The allocation decision factors price (from rate cards digitized in Shipsy), performance score, and carrier capacity in real time. A carrier whose composite dropped overnight loses booking share on the next request — automatically. This is what “performance enforcement” means. See rate card digitization + audit for how the price leg of this decision is built.
Tiered consequences. Shippers configure rules: score below 70 in any dimension for 7 rolling days triggers a pause on new bookings for that carrier on that lane. Score below 60 triggers an automated carrier notice. Score recovery above threshold automatically restores allocation. Humans set the thresholds; Astra enforces them.
Carrier-facing portal. Carriers see their own scores in a dedicated portal — the same numbers the shipper sees, same dimension breakdown, same event-level drill-down. This cuts the defensive back-and-forth during QBRs. If a carrier wants to dispute a score (e.g., “the geofence was wrong for that delivery”), they file a structured dispute that Vera adjudicates using the event trail.
QBR automation. Quarterly scorecards auto-generate from the same underlying data. Procurement leaders walk into carrier reviews with a ranked performance table, lane-level anomalies flagged, and a recommended allocation adjustment. Meeting preparation drops from days to minutes.
Early results
Shippers who move to live scorecard-driven allocation typically see: composite performance score across the carrier portfolio rise 8-15 points in the first two quarters (as laggards lose share and tighten up), FADR improvements of 3-7 percentage points network-wide, and a measurable reduction in “emergency” carrier swaps because Astra already rebalanced before things escalated. A global parcel leader spanning 65+ countries has used this pattern to drive progressive performance gains across their subcontracted last-mile network.
What’s next
Roadmap adds predictive scoring — weighting carriers not just on historical performance but on Astra’s forecast of their likely performance on a specific lane given current conditions (volume surges, weather, carrier-side capacity signals). See Astra’s planning agent workflow.