Preamble: I probably take Wikipedia too much for granted, and I know that server space, bandwidth, site administration, etc. don’t come for free.
But according to this appeal page…
How can it cost $10 million? Doesn’t look like a significant portion of that can be payroll, with only 35 staff. Where does the money go?
Well, they could easily blow a third of that on 35 staff members, not including the cost of any consultant or freelance work (and that’s assuming Jimmy Wales doesn’t care to pay himself like a proper CEO). I imagine lawyers eat up a lot of money, too. I can’t account for every penny in my head, but the $10M figure hardly shocks me.
It doesn’t cost $10 million.
The money they’re raising isn’t all going to Wikipedia. It’s going to The Wikimedia Foundation, which runs a dozen or more online things. Your donation is as much going to Wikiversity and Wikipsecies - I didn’t make those up - as it is Wikipedia.
Lots of info there.
Thanks ramel - that was what I was looking for - but my searches for ‘wikipedia statement of accounts’ etc. kept turning up wiki pages describing what that is in general.
I looked at:
http://upload.wikimedia.org/wikipedia/foundation/c/cc/FINAL_09_10From_KPMG.pdf
It looks like the expenses for Wikipedia are under projects. I don’t understand why the line item for operating when from 618,000 in 2009 to 2.8 million in 2010. Pages 9 and 10. I don’t know what that means and it is separate from salaries and internet hosting. It looks like rent was 280,000, so It isn’t clear where the other 2.5 million went.
what is Jimmy Wales’s incentive to be frugal? The more money a bureaucracy raises, the more it can expand and the more powerful will the guy in charge feel. Rothbard notes that traditionally there are strict guidelines for level of spending on various items in government bureaucracies but not in owner-run for-profit businesses because in a for-profit unwise spending would detract from the manager/owner’s income or equity whereas in governmental or “non profit” bureaucracy there is little downside to spending too much from management’s standpoint.
In the particular case of Jimmy Wales and Wikimedia I think that he has unwisely wasted the social capital and goodwill accumulated through the success of Wikipedia on spending on… whatever it is that he has already spent and is continuing to spend all the money on. Given how much he did for the public good with the Wikipedia project, if he were to have begun doing something else great, beneficial and actually functioning he would have been in a good position to solicit donations specifically for that next wonderful project. Instead he, AFAIK, did not come up with anything remotely useful, popular and “cool” ever since and instead is reduced to solicit money “to protect Wikipedia”. Protect from whom, one may ask 
Good points. I’m also not clear why ads are so firmly ruled out. If done properly, ads need not detract from Wikipedia at all, and in some cases, might add value to the site. It seems blinkered to simply rule them out altogether.
Come on, can you imagine opening Wikipedia and seeing a blatant banner ad emblazoned across the top, with some dude blatantly shilling for money?
Oh, wait . . .
I suspect it’s because the people paying for ads might want to have some control over what’s said (or not said) about them in Wikipedia – or at least it might look like this is happening. Wikipedia tries to be as objective and unbiased as possible, and carrying ads would make that more difficult.
Maybe, although I reckon a site with Wikipedia’s footfall and breadth of content should be able to find an advertiser more than willing to customise their ad stream for proper fit.
As far as I can tell, people generally tolerate (or even appreciate) internet ads as long as they:
[ul]
[li]are relevant and useful[/li][li]don’t mislead[/li][li]are distinct from, and don’t get in the way of the real page content[/li][li]don’t try to install malware[/li][li]don’t launch popups/overlays, play sounds, videos, or prompt for the installation of plugins[/li][/ul]
A big fish such as Wikipedia ought to be able to dictate terms such as these as absolute requirements, and still find an ad partner.
For one thing, Wikipedia is non-profit. It was specifically designed to be for the better of humanity, not some business venture. Yes, the founder is a Randian capitalist, but, honestly, he has very little to do with the site itself.
Plus, what you have outlined is a lot of work for a company to do that is doing fine with just donations.
Plus, who actually clicks on ads? Heck, this was the first place where I discovered computer literate people who didn’t run adblock software.
Doesn’t seem like a lot of work to me. Maybe a bit of a workup for the initial negotiations, but after that, a properly implemented ad scheme should more or less run itself.
Well, obviously some people do, or else the industry probably wouldn’t exist. I get about half a dozen clicks a day on the Google ads on my own site, and this brings in enough income to run it (obviously I’m not saying that’s necessarily scalable between a silly little site like mine and a huge one like Wikipedia).
Not everyone regards ads ad a menace. If they’re properly controlled and sensibly placed, I believe they can add value to a web page.
Firefox makes an amazing amount of money (something like $50 million per year) because of the Google search box on the default home page. Perhaps Wikipedia could do something similar, if it wants.
Thing is, there current model of asking for donations seems to be working pretty well. They just raised 16 million. I’m sure if they were going broke they’d be looking at ads and other things, but there isn’t really a reason right now.
I think it’s just the way the appeal was worded came across as really needy, to me.
Someone mentioned in another thread that they actually tested a bunch of different types of appeals, and went with the one that had the best response rate. So needy apparently works.
I think it’s more like 16 million, and that’s a surprisingly small number. They have to buy a lot of bandwidth and a lot of servers to handle the load (I’ll bet it’s in the hundreds). Then add in the switches, controllers, UPS’s, halon systems, air conditioning, security systems and other hardware that goes into a server farm installation. That means an IT staff, which means managers, which means HR and payroll, and tax accountants, and secretaries to answer the phone, and…
The electricity bill alone for a large server farm can be gigantic. I managed a small server farm for a while. It had around 20 servers in it, plus all the ancillary gear. It required a 10-ton air conditioner just to keep the room cool from all the waste heat. If the air conditioner failed, it was like being in an oven that someone just switched on. In fact, it takes about half the energy to keep a server room cool than it does to power all the servers.
And with the volume they have, it probably stresses the software infrastructure as well, which means they probably have a staff or programmers, bug fixers, and QA people.
The internet’s not free, and it’s not cheap. It may look that way if all you’re doing is hosting a personal web page on Squarespace or something, but once you get into real volume, you have to start paying real money.
Here’s Wikipedia’s Stats Page. According to that, last year Wikipedia’s servers processed 53,000 page requests per second, and average throughput of about 5 gb/s. That’s a huge amount of traffic. I’d guess that bandwidth alone costs the Wikimedia foundation somewhere north of a million bucks a year, but I have no idea what they really pay.
Frankly, the fact that Wikimedia can run their whole organization on that kind of money speaks well for their fiduciary care. I would have expected it to be much more. Google burns through money so fast that 16 million is a rounding error on their balance sheet. Google’s operating budget is over $10 billion per year. Their server farms have an electricity budget alone of over $2 million per month.
Given that Wikipedia is one of the biggest web sites in the world (top 5, I believe), 16 million per year is downright frugal.
Wikipedia doesn’t run its own data centers. Only a handful of the largest web sites do, precisely because it is so expensive - you have to have very specialized needs in order for running your own DC to make any sense. And a large global web site really needs multiple DCs, so increase your estimate by several times over.
I believe Wikipedia’s hosting costs listed in the links above are a couple of million a year. That’s about what I would have expected at a rough guess, for some sweetheart deals done with big DCs.
Ah, here are some details: