- Samuel's Blog - https://samuelgordonstewart.com -

Australia Post Statistics

One of the interesting things about Australia Post [1] is that it is possible to generate statistics based on forms they give you copies of, when you fill them in. This is made possible by the serial numbers applied to the services the forms are used for, being in numerical order.

One such form is the Registered Post [2] form, which gets filled in for every article sent via Registered Post, Australia Post’s insured post. Each Registered Post article is given a unique serial number, known as the “Registered Post Number”, this number is prefixed with the letters “RL” as is issued in numerical order. It is therefore possible to calculate how many Registered Post articles have been sent in a certain period of time based on these numbers.

On a recent check of the filing cabinet I found an old Registered Post form, from the year 2003, I also have two recent registered post forms.

Because I don’t have the exact times of each article being posted, and because having times would make the calculations too messy for quick use, some assumptions need to be made about the times of postage which do, unfortunately, add a small degree of error to the calculations. For the purposes of these calculations we will assume that each article was the last one posted on that day, which means that we need to subtract one item from each calculation to have “full day” figures.

For example, if item 15 was the last item sent on Monday and item 50 was the last item sent on Wednesday, a calculation of “50-15” would be a calculation of one item from Monday, and every item from Tuesday and Wednesday. A calculation of “50-16” would be a calculation of every item posted on Tuesday and Wednesday…this is effectively the same thing as subtracting one item from the “50-15” calculation.

Now that we have the theory behind us, we can perform some calculations:

The last figure does seem somewhat erroneous, and I suppose that it is possible that the numbering system isn’t entirely sequential, perhaps giving specific computers a block of numbers at a time. Regardless, the system is at least mostly sequential, and it would be unwise to expect perfect statistics from an external source anyway. Australia Post undoubtedly would have access to the true figures, and I would be surprised if they differ much from the ones above (with the exception of the last figure).

One interesting statistic from the Australia Post Annual Report for 2004-2005 [3] is the number of items posted, which for the 2003-2004 financial year was 5,307,500,000 and for the 2004-2005 financial year was 5,363,100,000, a total of 10,670,600,000 from the start of 1 July, 2003 to the end of 30 June, 2005, a period of 731 days, which makes an average of 14,597,264.02189 items posted every day (rounded to five decimal places).

If we use that average to estimate the total amount of mail posted from the start of 4 November, 2003 to the end of 10 March, 2006 (a period of 858 days), we can estimate that 12,524,452,531 (rounded to nearest whole number) articles were posted, which makes registered post (a total of 7,801,010 articles for the same period) a mere 0.06229% of Australia Post’s mail.

If one was willing to work out the number of days post offices were open during this period, it would also be possible to work out the average time between postage of each item, and each Registered Post item, sent through Australia Post.

I personally found that quite interesting, I hope you did too.

Samuel

Comments Disabled (Open | Close)

Comments Disabled To "Australia Post Statistics"

#1 Comment By thatsallfolks On March 14, 2006 @ 4:44 pm

I work for Australia Post, and the system you have used to calculate the number of registered articles sent is riddled with flaws.

The numbers used to identify each registered post item are not at all sequential – the labels are not computer generated but are distributed to Post Offices as stick-on labels on a large roll.

Each outlet has their own roll of uniquely numbered registered labels which are stuck on to the parcels after the registered post forms have been filled out. The numbers on these rolls are sequential for each postal outlet but are distributed at random.

Thus, it stands to reason that a post office which does little business may take a year to use the entire roll of labels, while busier post offices could use several in a month.

The roll of labels we are currently using follow the sequence
RL15 etc and we are receiving parcels with labels from anything from RL13 to RL38.

No accurate conclusions can be made based on the data and formula you have used.

#2 Comment By thatsallfolks On March 14, 2006 @ 8:35 pm

The reason the last figure was so inaccurate is because those numbers were specific only to the Australia Post outlet you registered the articles at. Hence, at that particular outlet 33 items were registered over that 4 day period.

#3 Comment By Samuel On March 14, 2006 @ 9:07 pm

Ah, OK then, thanks for letting me know thatsallfolks. It was a fun bit of analysis anyway.