Original research · August 2026
What 224,604 parcels show about how mailrooms actually run
Two years of measured data from sixteen mailrooms. The intake peak is not when everybody plans for it, three carriers bring four fifths of everything, and most of a day's work lands before 2pm.
- 224,604parcels measured
- 24months of data
- 16mailrooms
- Zerosurveys, estimates or guesses
The intake peak is October, not September
Every student building in the country staffs up for September. The parcels arrive a month later.
Index against an average month, where 100 is average. Eight student accommodation sites, 209,686 items, 24 months.
October runs at 178% of an average month. September manages 113%. October is 5.8 times the volume of August, which is the quietest month of the year by a distance.
It makes sense once you see it. Move-in week generates footfall, not deliveries. The parcels come once people have unpacked, found out what they forgot, and ordered it. September and October together take 24% of the year's parcels in two months, against the 17% an even spread would give them.
Amazon is 38% of everything
Share of 210,655 items.
Amazon, Royal Mail and Evri carry 79% of everything that arrives.
Nearly half the day has arrived by 2pm
-
49% by 2pm
Just under half of a day's parcels are already in the building by two in the afternoon, and 62% by three.
-
1pm is the busiest hour
16.6% of the entire day arrives in that single hour, more than three times the 10am hour.
-
Monday is the heaviest day
18.8% of the week against Sunday's 6.2%. Weekend orders land together.
Cover is usually spread evenly across opening hours. The arrivals are not: they are concentrated into the middle of the day, which is also when reception is busiest with everything else.
Most parcels are still there four hours later
62% of parcels are still in the mailroom four hours after they arrive.
The median item is collected in six to eight hours, so the room does empty. It empties because collection is chased rather than waited for: uncollected items are reminded automatically on a schedule, and what would otherwise be a growing pile becomes a queue that clears.
The outcome is 98.4% collected. Of 209,686 items more than a week old, 3,245 were never picked up. In buildings whose entire population changes every September, 1.6% is a remarkably small residue.
The gap between "62% still here after four hours" and "98.4% eventually collected" is the whole argument for automated reminders.
The Data
Source. Sixteen production installations, 1 August 2024 to 31 July 2026. 224,604 individual items. Counts and aggregates only.
The cohort. Seasonality, carrier and timing figures use the eight purpose-built student accommodation sites in London with complete records, totalling 209,686 items
Turnaround. Measured as the interval between an item being scanned on arrival and being handed over.
Figures may be cited with attribution and a link to this page. Rounding is always down.
Questions
Where does this data come from?
Sixteen production Traizr databases covering 1 August 2024 to 31 July 2026: 224,604 individual items. Counts and aggregates only, no personal data. The seasonal and timing figures use the eight student accommodation sites with complete records, 209,686 items.
Why is October busier than September?
Residents arrive in September and order in October. Move-in generates footfall rather than parcels; the deliveries follow once people have unpacked and worked out what they forgot. October runs 178% of an average month against September's 113%.
Is this representative of all student accommodation?
It is eight purpose-built sites in London over two years. That is a real sample rather than a survey, but it is London PBSA and not a national picture. The seasonal shape should hold anywhere the academic year starts in September; the carrier mix will differ by country.
Can I cite these figures?
Yes, with attribution to Traizr and a link to this page. If you need a cut of the data that is not published here, ask and we will see whether it can be produced without identifying a customer.