Penny for your thoughts

The Wikimedia Foundation is in the midst of a two-month long fundraiser, running from October 22 to December 22. Direct fundraising from the online community provides the vast majority (96%) of Wikimedia’s support. Given Wikipedia’s profound reach, it’s not an inexpensive project to run – the budget for Wikimedia for 2007-8 is $4.6 million, with approximately $2.6 million going towards the servers and bandwidth neccesary to keep the project running.

(Some quick disclaimers, before we get into the meat of this post. I’m on the Advisory Board of the Wikimedia Foundation, and am on the mailing list of the Fundraising committee, though I’ve been a very quiet participant, as I know a lot about fundraising from foundations and basically nothing about direct donations. That said, there are no secrets in this post – I’m linking to publicly available statistics. And yes, I’ve made a donation to the Wikimedia campaign, and you should too.)

In the past 17 days, the Foundation has raised over $478,000 in online gifts. That’s a pretty amazing number, on the one hand, and a concerning one, on the other hand. If Global Voices could raise that much money online in a month, I’d be out of a job, as our annual budget is not much higher than that sum, and I spend far too much of my time convincing generous individuals, corporations and foundations to support our efforts.

On the other hand, it’s significantly less money than my local public radio station raised in its last fundraising drive: $801,000. Those drives are less than a week long, and there are three a year, raising a large percentage of the station’s annual income. Wikipedia reports the station’s listenerbase at 400,000 – 5,874 (1.4%) of those listeners gave $136 each on average in the most recent drive alone.

What’s Wikimedia’s audience? Alexa estimates that between 8 and 9% of all Internet users access Wikipedia.org the site in the course of a single day. Let’s ignore the vast majority of the internet and focus on the 212 million US internet users, most of whom have credit cards and many of whom have disposeable income. If 1.4% of them contributed to Wikimedia, that would be 2.96 million supporters of the project. If they were as generous as WAMC listeners, that would be over $400 million. If they gave the amount the average Wikimedia donor gives – $28.03 – we’re talking about $83.2 million.

If it’s true that 9% of the world’s internet users visit Wikipedia in a day (and it’s probably not, as Alexa’s figures are based on extrapolation from a fairly small sample set which isn’t perfectly geographically distributed), that’s roughly 112 million visitors, 99.999983% of whom don’t donate. (Still haven’t donated? Assuage your guilt here.) Some of the people who don’t donate probably can’t, either because they lack the funds or because they lack the credit card or checking account to make the gift. But some very large number of people is choosing not to support the project despite visiting the site. Why?

One possibility is that internet users are cheap. This seems to be the conclusion that many commentators are drawing from Radiohead’s experiment with allowing fans to choose the price they’d like to pay for the band’s latest opus. 38% paid between $0.01 and over $200 for the ten mp3s, while 62% chose to pay nothing at all. The average price paid was $6, significantly below what a conventional CD would cost… Other artists have had better luck with downloadable albums with adjustable pricing – avant-folk artist Jane Siberry has offered music from her website for the last couple of years, suggesting prices per track, but inviting users to pay more, less or accept the track as “a gift from the artist”. Only 19% of her listeners chose this option – 79% of those who paid gave the asking price, and 15% chose to pay more than the asking price.

One possible explanation for the difference in Siberry and Radiohead’s experiences is that one prompted a price, while the other didn’t. The average price paid for Radiohead’s album may reflect what people think a CD should cost, even if the music industry would like it to be three times as much – perhaps paying $6 is a type of protest, asking for fairer pricing in music.

Another possibility is that a lot more people heard about Radiohead’s policy than about Siberry’s and visited their site to find free music. Users coming to Siberry’s site are already fans and might be more likely to pay than listeners who don’t know Radiohead but know they can get a new album for free.

Looking Wikimedia’s fundraising from two years ago suggests that there might be something to this theory. Two years ago, Wikimedia raised almost $400,000 in 21 days… pretty close to the results the current fund drive is having, despite the fact that Wikipedia’s reach is vastly larger now than it was then. (Alexa suggests that roughly 2% of a slightly smaller internet visited Wikipedia by the end of 2005. With 1.01 billion net users in December 2005 and 2% daily share, that’s 20.2 million visitors, versus 112 million.) It’s possible that Wikimedia is supported by a small core of users who donate to support the site, and that expanding the userbase of a site won’t necessarily expand the donor base.

This shouldn’t be a major problem for ad supported sites. As a site gets more traffic, it shows more ad banners and, one presumes, gets more clicks and therefore more revenue. But Wikimedia doesn’t sell advertising. As more people use the resource, bandwidth and server bills go up, and free riders become a real problem.

When public radio stations in the US run fund drives, they spend a good deal of their airtime reminding listeners of their economic model. Despite the name “public radio”, most of a station’s funding comes from listener giving and local underwriting – part of getting listeners to give is reminding (haranguing, nagging) them of their responsibility to the station to ensure its continued survival.

The internet, on the other hand, appears to be moving inexorably towards a model where everything is free and ad supported. My friend danah boyd, talking about teens use of social networking software, noted that most users are apparently not turned off by ads on MySpace or Facebook. On the contrary, they assume “If it’s got ads on it, it will be free forever!” In a world where nearly everything is free due to ad support, it’s quite likely that a casual user may not know that Wikipedia and other Wikimedia projects are supported almost exclusively by user donation, and a banner on top of pages may not be sufficient to challenge the paradigm that everything on the internet is free.

In fact, charitable giving is now free as well. Sort of.

A new site called FreeRice is generating a lot of attention from web users and has been growing quickly. The site invites you to participate in a multiple-choice vocabulary quiz and awards you ten grains of rice for each word you get correct. The rice is given to the World Food Programme, which gives it to people in need. You don’t pay with anything but your attention – the money for the rice is given by the advertisers whose banners feature on each page of the site.

FreeRice claims 73,566,480 grains of rice donated so far today… which sounds like a lot, until you realize that there are between 35,000 and 50,000 grains of rice in a kilogram. 1471 kilograms of rice donated in a day doesn’t sound like nearly as much. Especially when you discover that prevailing world prices for rice run at about $12 a hundredweight, or 45.5 kilos. 73 million grains might run about $387.

To donate 73 million grains of rice, FreeRice is serving roughly 10 million pages (10 grains for a successful answer, and some percentage of answers are going to be incorrect), most with three ads on them. perigee on LiveJournal takes a crack at estimating revenue from FreeRice and argues that the site profits between $4.44 and $17.99 for every 150 page impressions. I think his estimates are hugely optimistic in terms of what ads sell for these days – he suggests a cost per 1000 ads (CPM) of between $10 and $40. We rarely were able to sell ads for that much in the good old days (1998), and now advertisers have wised up and are only paying for clicks on ads. If 100 page impressions equals 1000 grains of rice (I’m good at vocabulary), that’s about half a cent donated to the World Food Programme. If I click on one ad, I might make somewhere between five and twenty cents for the site administrator.

Did I mention that FreeRice.com isn’t a charity? It’s the latest project of John Breen, the programmer who created The Hunger Site, a for-profit company that encouraged visitors to click on a button on the site to donate food to charity – it too is ad supported. The Hunger Site is no longer run by Breen, but by CharityUSA, a for-profit company that runs over a dozen “click to give” sites.

There’s nothing wrong with running for-profit websites, and certainly nothing wrong with for-profit companies giving to charity. But FreeRice is far from transparent about what percentage of its money gets given to charity and what percent supports its operation, or is revenue for the site’s creator – perigee has asked Mr. Breen to go on the record on these issues.

My point isn’t to trash FreeRice or the people who feel good about giving to it, though I certainly think there are better models for projects like this one. It certainly wouldn’t be hard for FreeRice to include links to stories about hunger and food security if their concern is getting people to understand more about the issue. Or as Hannah Farber at jspot points out, they could just urge people to give real money directly to charities that work on these issues.

No, what interests me about FreeRice is the reinforcement it gives to Internet users that their attention is a valuable currency. In an economy where all anyone wants is a moment of your attention, it’s possible for many things to be free, so long as they’re sufficiently popular, and so long as their creators are comfortable monetizing them by being willing to share their attention with an advertiser. Wikimedia, god bless them, isn’t playing the game, and they may find themselves in an increasingly difficult spot as they reach more and more users who don’t understand that they’re supposed to help pay for the content, not just use it.

My friend Chad Orzel has a useful reaction to this post over at his excellent blog, Uncertain Principles.