OddThinking

A blog for odd things and odd thoughts.

Courier versus Australia Post – Reliability Review

I need to send a document that is valuable to me interstate.

It isn’t urgent (say, within a week or two). However, it is important that it not get lost, as it will (allegedly) cost many hundreds of dollars to replace.

I asked on Twitter and Facebook whether I should send it via Australia Post or via a courier. I didn’t get a satisfactory answer, so I did the research myself. This is to help other people concerned about the reliability of Australia Post.

For the sake of calculations, I am assuming that the item is worth $1000. If it was worth more than $10,000, my decision would need to be revisited.

The alternatives I am considering are:

  • Australia Post – Regular Letter
  • Australia Post – Express Post
  • Australia Post – Registered Letter
  • Australia Post – Registered Letter with Insurance
  • Independent Courier, such as FedEx, DHL, Courier’s Please, etc.
The Express Post “guarantee” is merely to provide a replacement Express Post envelope if the letter is late. The guarantee doesn’t really help here, and is not factored into the equations.

I have limited information available to me, but what I do have turns out to be enough to make a decision.

Prices [Ref 1, 2]:

  • Regular Letter: $0.55
  • Express Post: $4.50
  • Registered Letter: $3.40 (includes $100 insurance)
  • Insured Letter: $3.40 + $1.30 per $100 insurance (first $100 free) = $15.10
  • Courier: Around $52 (Sample: DHL, minimum weight = 0.5 kgs)

There is a big premium for using a courier. They pick up from your door, which is good if you can be sure someone will be there, and the address isn’t impossible to describe, but I would rather drop it off at a post-office.

It might be worth it, though, if courier are more reliable.

Alas, I don’t have figures for courier reliability.

I do have some for Australia Post though! According to their annual report (p129), 95.5% of letters are delivered on time, and 98.8% are delivered within a day of the expected time.

This leaves 1.2% unaccounted for. Some of them may have been delivered later still – perhaps within the “week or two” deadline I set. Some may have been delivered to the sender. Perhap the proportion that were lost is much smaller.

Furthermore, I can hope to do better than the average by taking a few simple precautions – making sure I get the destination address right, having a relatively straightforward destination address, writing clearly and ensuring there is sufficient postage.

Nonetheless, I am going to assume that all 1.2% are lost for this next calculation.

Similarly, I would expect that the Express Post and Registered Mail have higher reliability figures, due to the additional scrutiny, but I have no support for that claim, so I am treating them as regular letters for this next calculation.

If we successfully send the $1000 letter 98.8% of the time, and lose it entirely 1.2% of the time, we have an “expected value” (in the statistical sense) of $988. We lose $12 in value through the unreliability. (Realistic figures would be reduce that $12.)

That is a very useful figure, as it gives an upper maximum of how much we are willing to spend to further protect the asset. If we spend more than $12, we will be losing out. (That is, for every 1000 letters we send, we will have less money and assets at the end.)

So, we can rule out couriers. Even if their price was a quarter of the quote that I received, and their reliability was 100%, it still wouldn’t be good value.

The next question is whether a registered letter is better than a normal letter. The $100 insurance is worth $1.20 (1.2% chance of receiving $100), so the premium we are paying is $3.40 – $0.55 – $1.20 = $1.65. That requires the receiver to sign a card to receive the letter, and that card is then sent back to the sender. It provides a level of non-repudiation; the receiver can’t deny that the letter was in their control.

Is that worthwhile? That depends on how frequently the recipient loses their mail and blames it on the Post Office! It also is probably of great assistance if the Post Office does lose the mail – it proves (?) that you sent it. In my case, I have decided that this is worthwhile, but not based on clear facts.

Which leaves the question of insurance – each additional $100 insurance is worth, at most, $1.20 to us, but costs $1.30. Therefore, we should not accept the insurance.

As always, with insurance, it is a poor bet from an expected value perspective. If this letter represents your last $1000, it probably has more utility to you than 1000 times the marginal utility of $1. In which case, this argument wouldn’t hold. Similarly, you might not use this argument when evaluating home insurance. It would be even more inappropriate for evaluating health insurance or public liability insurance, because your costs might be more than your current assets.

In conclusion, I haven’t proven that Australia post is more reliable than couriers, or even equally reliable. I haven’t come up with a particularly accurate estimate of the reliability of a clearly and correctly-addressed piece of registered mail. However, I have come up with an upper-bound to how much you should be willing to pay for higher reliability and for insurance.

From that, I can conclude uninsured mail (perhaps registered) is a better option than insured mail or couriers that charge significantly more than Australia Post.

For previous discussions about economics and Australia Post, see Separation of Church and Postage.

Comments

  1. You have come up with an upper bound for how much you should’ve been willing to pay during the time that the survey was taken. For all you know since then 50% of all mail is lost…

  2. Sunny,

    The survey covered each quarter from July 2008 – June 2009.

    I don’t have any reason to believe Australia Post’s handling of mail has degraded significantly since July 2009.

    I haven’t studied the details, but I haven’t any reason to believe that the sampling or analysis was so substandard so as to invalidate the coarse measures I was using. It had an independent auditor.

    So, I agree that such a sampling problem is theoretically possible, I don’t see it being a problem here.

    Update: Corrected typo. “It” had an independent auditor, not “I” had an independent auditor! 🙂

  3. Australia Post have lost three parcels, couriers that I have used have lost 0 parcels. To me couriers are more reliable. Also their customer service is far better

    Richard

  4. I am sorry to hear about your three lost parcels, Richard, but your self-selected anecdotal data is hardly sufficient to affect my analysis, especially without data as to how many parcels you have successfully sent through each service. In any case, I am talking about letters, not parcels, which I assume have different success rates.

  5. One thing is for certain, whenever you rely on an Express Post to be delivered on time or at all – sure as can be it vanishes form the face of the Earth – I’ve had two go missing in a week! No trace of them!

  6. Shaun, oh dear! Two examples in a week is clear proof that all Express Post parcels disappear “for certain”, and not, say, that you have experienced an anecdotal 1/7000 chance of being in the 1.2% twice in a row.

  7. You are missing the point.

    Australia Post offers express freight OR registered and insured. Not both at once.

    Couriers, offer both at once.

    I think, the 1.2% is wrong. Its way higher than that, they are fudging the stats. Not hard to do. Just only include those parcels that AU Post lost for certain. Ignore the ones where they just left it on the front porch, and someone took it(happened to me) or those where the package was delivered, opened and empty(again, happened to me).

Leave a comment

You must be logged in to post a comment.