Peak Season, Planning May 7, 2026 · 6 min read
Forecasting, Freight
May 3, 2026 · 7 min read · Operations desk
Air forecasts fail for one reason: they predict orders, not shipments. Orders bunch, split and shift service levels. Forecast at the shipment level — by route, by week, by service level — and accuracy jumps immediately.
Blend three inputs: sales' commercial forecast, last year's shipped pattern, and known events (promos, launches, plant shutdowns). Weight recent weeks heaviest; freight demand has a short memory.
Finally, forecast the range, not the point. A route that needs "four tonnes, maybe six" gets a very different plan than a flat "five". Share the range with carriers and you'll find uplift far more forgiving.
Reconcile every Monday. Compare what shipped against what was forecast, note the miss by lane, and carry the lesson into next week's numbers. A forecast reviewed weekly improves fast; one reviewed quarterly never does.
Share the range early with capacity owners. Carriers can protect uplift against "four to six" far more readily than they can rescue a single fixed number that turns out wrong on Friday afternoon.
Forecast the range, not the point
A lane that needs "four tonnes, maybe six" gets block-space plus a named backup flight. A flat "five" gets neither — and fails exactly when it matters.
A forecast that works
Three inputs, one weekly rhythm
- Blend the commercial forecast, last year's shipped pattern and known events.
- Forecast by route, by week and by service level — never orders alone.
- Publish the range to carriers and reconcile the miss every Monday.
Keep reading
More from the operations desk
Peak Season, Planning 
Reporting, Operations May 5, 2026 · 5 min read
What growing air cargo teams need from operational reporting
Read Full Article
Partner with an air freight team committed to speed, security, and precision.
Tell us about your routes, volumes and timelines — we reply within one business day with a plan and transparent pricing.