Download the statistics page's KPIs as rows of metric, period, dimension and value

Use this to chart the same KPIs the Statistics page shows, without rebuilding them: shipments per month, by carrier, lane and country, arrival punctuality, port dwell percentiles, on-time percentage, CO2 and produce. Metrics, with the dimension each one uses: shipments_created (monthly, transport_type: ocean or air); shipments_by_carrier (carrier), shipments_by_lane (lane), shipments_by_origin_country and shipments_by_destination_country (the country): shipments in the window; arrivals_on_time, arrivals_late_up_to_5_days and arrivals_late_over_5_days (by carrier and by lane): arrivals against the original ETA; port_dwell_median_days, port_dwell_p10_days, port_dwell_p90_days, port_dwell_q1_days, port_dwell_q3_days and port_dwell_containers (destination_port): days containers sit in the destination port; on_time_arrival_pct and arrivals_measured (monthly, no dimension); co2_tonnes (monthly, no dimension); shipments_by_produce (monthly, produce); shipments_by_supplier (monthly, supplier); incidents (monthly, incident_type: delay, rollover, free_time_breach, document_error or customs_hold, unit incidents); and for each of carrier, forwarder and supplier, over the window, incidents_by_<dimension> (unit incidents), shipments_with_incidents_by_<dimension> (unit shipments) and incident_rate_by_<dimension> (unit per_100_shipments). Incidents are derived from delays, rollovers, free time breaches, document corrections and customs holds, counted in the month their shipment was created. Ranked metrics keep their top entries as the page does (10 carriers, 10 lanes, 10 ports, 8 produce, 8 suppliers, 10 carriers, forwarders and suppliers by incidents); use the shipments feed for the full list. Append .csv or .json. Every period is the month a shipment was created in, not the month it sailed or arrived, and the figures include archived and orphaned shipments, exactly as the Statistics page does.