[WikiEN-l] Real-time mirrors as a net revenue _source_?

Mon Apr 10 09:59:32 UTC 2006

Ryan Delaney wrote:
> On 4/9/06, Tim Starling <t.starling at physics.unimelb.edu.au> wrote:
> 
>>
>>That's arguable. You could say that there are many copies of Wikipedia on
>>the Internet, and only one
>>of them is ad-free, that is the one at wikipedia.org. Most readers come to
>>Wikipedia via the search
>>engines, and a large part of the remaining 80% of advertising revenue will
>>go towards spamming those
>>search engines with irrelevant keyword-pumped advertising-laden copies of
>>Wikipedia pages.
> 
> 
> 
> Okay, but isn't that the situation we are already in? The only difference is
> that the people running the advertising-laden copies will have to pay for
> the privilege of doing so.

It's the difference between harrassing and encouraging. What we'd be selling would be the ability to
set up a Wikipedia mirror for $10 per month plus a couple of hours of setup time, with the security
of knowing your service won't be randomly cut off by the Wikimedia server admins. Currently you
either have to buy more expensive hosting with enough disk space to hold a complete copy, or you
have to evade the blocks and remote load from our servers.

Of course, if the venture is successful and the site becomes popular, then the costs will increase
in proportion to the traffic, but so will the revenue. That's why the startup costs are the only
financial risk involved. If you reduce or eliminate them, then you encourage the proliferation of
mirrors.

How many mirrors is enough? 1000? 10,000? 100,000? Eventually I imagine the market will become
saturated, when a new mirror can't recover its minimal startup costs, even with the most aggressive
SEO techniques. What will the Internet look like then? Will the average user be able to find
independent information in the search engines, which didn't come from Wikpedia?

-- Tim Starling