Trade Me Math 101

January 30, 2011

Someone smart once told me that the best way to determine a price for your product or service is to find the point at which the customer will wince, and then take half a step backwards.

I was reminded of that this morning reading some of the comments in the media and online about the latest round of Trade Me price increases announced earlier this week.

The official justification is based on a 91% increase in traffic in the two years since the last increase (holy crap!)

But, I wonder this tweet from David Slack perhaps describes it more succinctly:

That Trade Me price rise rationale in full: “Because we can.

Perhaps not surprisingly, this announcement was enthustically reported by the NZ Herald, but as far as I can tell ignored by Fairfax.

The Herald article is terrible – as well as being a shameless plug for Sella, which they partly own (although that isn’t disclosed in the article), they also let themselves down with some shoddy maths by miscalculating the impact of the increase on a hypothetical $1000 item.

Choosing $1000 as the example is in itself not very smart – most stuff that sells on Trade Me is much cheaper than that.

So, as a service to numerically-challenged or commercially-conflicted reporters, here is a quick back-of-an-envelope analysis of the impact of this change:

Firstly, let’s take a random sample of 100 listings – auction #348,000,000 through #348,000,099.  These were all listed on the site earlier this month and closed over the last week or so.

Of these, 24 sold and attracted a success fee.  Six were “Buy Now” sales, another three were “Fixed Price Offer” sales and the remaining 15 were standard auctions.  The sale prices ranged from $1.50 to $1300.

Obviously this is way too small to be a statistically relevant sample, but actually those values and proportions feel about right to me, based on the sale rates that are published on the site and a quick look at the list of items closing soon.

The total success fees charged on these 24 sales was $141.98.  If the new fee structure was in place that amount would have been $151.53, which represents an increase of just over 6.7% (the Google Spreadsheet linked to above has the full calculation).

To really understand the possible impact of this change though, you need to multiply that increase across the hundreds of thousands of auctions that close on the site each day.  Trade Me don’t actually publish those numbers, but with a bit of digging you can roughly work it out.

As I type, auction #349,438,751 is about to close (I’m going to assume that this was listed this time last Sunday evening), and the latest auction is #351,145,747. That would imply that in the last week there have been ~1.7million new listings (about the same number currently listed on the site).

Extrapolating the success fee revenue we calculated above across this number of listings gives revenue of $2.576m per week soon vs $2.413m per week currently.  So, approximately $163,000 extra per week or nearly $8.5m per year.  That should pay for some good times!

Of course, all of this assumes that the fee increases have no impact on the number of listings on the site, which may or may not turn out to be true.  I’d be confident though that the smart people at Trade Me will be watching this closely after the change is made to give themselves confidence that the overall impact is positive from their perspective.  I doubt they will be paying much attention to the wailing on the message boards, or in the NZ Herald comments, as similar threats, which have accompanied every single fee increase announcement since the very early days, have to date failed to make any dent in the continued growth in popularity of the site.

Anyway, the much more interesting question, which I haven’t seen anybody ask, is this: what good news does Trade Me have up their sleeve to announce on the 7th Feb? 🙂

Disclosure: I am a Trade Me alumnus, but I left when the site was just a fraction of its current size, so really, what do I know?!