[Rudolph Hucker]

Just a thought about your numbers: evilBay 'items offered' includes a large number of unsnipeable items such as all the Motoring Classifieds and all the Buy It Now items with no auction and all the stuff in EvilBay shops.

Regards,
Rudolph

____________________________________________
You do not stop playing because you get old, you get old because you stop playing

[the3coopers]

Actually I am amazed that the numbers are so small !

Consider - this is an international service... I only buy on ebay.com.au, and currently there are 1,336,432 items listed on ebay Australia.

Note that this doesn't include Stores, however it does include things that are listed overseas but will ship to Australia. Let's say that there are 1 million true Aussie items. Now, assuming that ebay internationally is 16 million, and that there are 172,000 automated snipes - that means about 11,000 auto-snipes for Australia.

I am very glad that the usage of sniper services is very limited - it means I have a much better chance :-)

[cross3_16]

I like your Fermi calculations Mario.

[mario]

Few days ago, Gixen passed the mark of 10,000 users. While this number is by no means small, Gixen is still much smaller than the biggest players in the sniping arena.

Seeing some numbers and estimates available at auctionsoftwarereview.com and auctionbytes.com, I decided to publish my own analysis, that will probably be much closer to the true numbers.

Methodology: I use Gixen own numbers and traffic reports from alexa.com to estimate numbers of Gixen competitors. While this methodology is still prone to some error, I believe it is far more reliable than throw-dart methodology and self-reported figures.

So here it goes, some Gixen numbers first:

Overall number of users: 10,200
Scheduled snipes per week: 9,100
Auctions won in a single week: 3700
Active users in a given week: 2300

Estimating overall numbers for long-running services would be very difficult, as some of the biggest sniping services have been around for a long time now and may have many inactive users. This is why I will focus on weekly number of placed snipes only. Based on these numbers, and traffic reports from alexa.com, I estimate the following numbers:

Auctionsniper.com: 6 times Gixen size (or 54,000 per week)
Auctionstealer.com: 2 times Gixen size (or 18,000 per week)
Auctionstealer.com (through auctionlotwatch.co.uk): 1 times Gixen size (or 9,000 per week)
eSnipe.com: 2 times Gixen size (or 18,000 per week)
bidnapper.com: 1.5 times Gixen size (or 13,500 per week)
justsnipe.com: around the same as Gixen (or 9,000 per week)
snipeswipe.com: 0.5 Gixen size (or 4,500 per week)
ezsniper.com: around the same as Gixen (or 9,000 per week)
powersnipe.com: around the same as Gixen (or 9,000 per week)
snip.pl: 2 times Gixen size (or 18,000 per week)

There are a number of smaller services: bidrobot.com, stealthbid, bidslammer, esteal, and others, but they all together probably have 10,000 per week at most. When you add all these, you get a number of around 172,000 snipes placed per week by automated services, 68,000 of which are winning. This does not include snipes placed by user-installed software or manual sniping.

Now let's estimate the percentage of auctions being sniped. Gixen gets outsniped in around 3% of all auctions, and since it places its snipes at around half-sniping-time (5-6 seconds), my guess is that around 6% of auctions are being sniped. This is expected, and in line with figures reported by other auction sites.

As auctionsoftwarereview.com reports in ebay stats (the report may not be recent), at any given point ebay has around 16 million listed items. It is fair, then, to assume, this is the number of auctions that are listed weekly, as most auctions are listed as 7-day. This means that automated sniping services snipe around 1% of all the listed auctions, while the remaining 5% are sotware-based or manual. This result is a bit surprising, as automated sniping is the most reliable method. One can try to explain it with lack of trust, and/or high pricing of some of the services.