When Do Delivery Drivers Actually Work? Data from 140,000 Routes
Short answer: Most planned delivery routes start on a weekday. Monday to Friday account for 87.3% of the 139,861 routes with a planned start day, with Tuesday the busiest at 18.2%. Weekends carry 12.7%, and Sunday is the quietest day at 3.8%. Among routes with at least one stop, the median is 4 stops, so the typical route is short.
Key findings
139,866 routes have been planned on the Geo2 platform since January 2024. 63,845 were planned in the last 12 months, and 73,931 are completed.
Weekdays carry 87.3% of planned routes. Tuesday is the peak day at 18.2%, and Friday is the lowest weekday at 16.5%.
Saturday is 8.9% of routes and Sunday is 3.8%.
Among routes with at least one stop, the average is 8.2 stops and the median is 4. One route in ten has 18 or more stops.
The average route carries 7.2 consignments.
740,547 proofs of delivery have been captured, and about 14,900 users have planned at least one route.
Which days do delivery routes run on?
The planned start days show a clear weekday pattern. Monday to Thursday each take between 17.1% and 18.2% of routes, and Friday dips to 16.5%. Weekends drop sharply from there.
What does the weekday pattern mean for drivers?
Saturday alone is more than double Sunday, at 12,408 routes against 5,371. For a driver on a five-day rota, the weekend is where routes thin out, and that is the rest period the schedule gives back. For a driver who wants extra hours, the same gap is the opportunity: Saturday still carries about one route in eleven, and it is less crowded in planning terms than any weekday.
For dispatchers, the peak is Tuesday, not Monday. Staffing and vehicle plans built around a Monday peak will miss the busiest day of the week. Planning the week's routes together, rather than one day at a time, is what delivery route planning and optimization is designed for.
How big are delivery routes, and where does optimisation matter most?
Most routes are short. Among routes with at least one stop, the median is 4 stops and the average is 8.2, because a minority of long routes pulls the average up. One route in ten has 18 or more stops. The average route carries 7.2 consignments, so a single route often covers several customers' orders at once.
The reasoning here is about where sequencing pays off. A four-stop route leaves little room for a better order. A route with 18 stops or more has far more possible orders, so the gap between an order picked by hand and an optimised one is larger. That makes optimisation most valuable on multi-drop routes, where a single run covers many drops.
A useful check for any dispatcher is to split their own routes into three bands: under 5 stops, 5 to 17 stops, and 18 or more. If most of the work sits in the first band, sequencing is a small gain and the bigger wins are elsewhere, such as address quality. If a meaningful share sits in the 18-plus band, the order of stops matters a great deal.
Is proof of delivery standard practice?
The platform has captured 740,547 proofs of delivery. That is a count of captured proofs, not a claim that every route used proof of delivery at every stop. What it does show is that proof capture is a regular part of how drivers and dispatchers use the platform. For how Geo2 captures it at the door, see electronic proof of delivery.
For drivers, proof capture is part of the stop rather than a task for the end of the round. A skipped proof leaves a gap that is hard to explain later, so it is worth building into the routine from the first delivery.
What this data does not tell you about driver hours
A planned route is a unit of work, not a working day. One driver can run more than one route in a day, so these counts say nothing about how many hours anyone worked. The weekday split uses planned start days, so it shows when work was scheduled, not when every driver was on the road.
The data also covers planned routes, not only completed ones. Of the 139,866 routes planned, 73,931 are marked completed. This data does not explain the difference, so do not read the completed figure as a measure of how much work was done.
How was this data counted?
Source: Geo2 platform data from the mobile app and dispatch, January 2024 to October 2026.
What was counted: planned routes, planned stops, consignments, proofs of delivery and users who planned at least one route. Weekday shares use each route's planned start day.
Stop basis: the average, median and 18-plus share of stops per route use only routes with at least one stop, counted from the stops records. They are not calculated from the loads count or divided by all planned routes.
Exclusions: deleted routes are not counted. Five routes have no planned start day and are excluded from the weekday shares only.
Coverage: the whole Geo2 platform rather than a sample. It describes Geo2 users only and is not a survey of the wider delivery industry.
Not reported: time of day, distance, duration and emissions. The data for these is not reliable enough to publish.