According to the below Times article (Dec 2007 - a month later), 62% paid zero and 12% accounted for more than 52% of revenue. Average price paid was $6 globally. US average price was $8.05; outside the US was $4.64.

http://technology.timesonline.co.uk/tol/news/tech_and_web/th...

Article also quotes Fred Wilson: "This shows pretty conclusively that the majority of music consumers feel that digital recorded music should be free and is not worth paying for."

As for me, I paid zero. And then bought the vinyl a few months later.

This article is ignorant and extremely biased. The first paragraph totally ignores the fact that consumer behavior radically changes when the price of a good hits zero. People who would otherwise not listen to Radiohead downloaded it -- just because it cost them nothing. The assumption made is that everyone who downloaded it would have ran out and bought it in stores, which is a terribly flawed premise. Did they listen to it once, or did they listen to it a thousand times? They may or may not have bought it, but it's impossible to tell.

I feel as though they did fantastic, considering people had the choice to get the exact same good for free or not free, and an amazing 38% of them paid for it. I'd feel safe betting that if other artists could pull this level of revenue off, they'd be beating the deep cuts taken by the record industry.

The data from Radiohead's experiment is pretty interesting. There is one thing that they did poorly though. Consider the person who downloads the album for $0.00 because they want to hear the music before deciding how much they are willing to pay for it. Radiohead's website did not allow that person to go back and give money after the initial download. The only way to pay any money in that case, was to download the album a 2nd time using a 2nd email address (couldn't download twice with the same email) and include a $$ value for the 2nd download.

Credit me with 2 downloads, one for $0.00 and one for $5.00. If a lot of people did that, it would inflate the total number of downloads and reduce the average paid. I wonder how many people did something similar?

this new website has a problem, too.

    I'm afraid you card payment was not sucessful.
    You may wish to try again or contact us with the details below.
    OrderNumber TKOL1-345345345345345345353
    Error details: 7: Transaction declined..

i tried twice, then got an email saying

     If you have placed more than one order in error, please be assured your email has now  been logged and the refund will be made over the next few days.

It's worth noting that an average price of $6 is a fantastic result for the band. I don't know what their overheads were for the digital distribution, but they were almost certainly much less than what they'd have paid a major label for a traditional CD release, where bands are reckoned to get about 10% of the profits. 10% of $12 is much less than 50% (say) of $6.

Picking nits about the release of sales data: The figure being cited was provided by a comScore, based on independent tracking data. The last line of that article states "A spokesman for Radiohead was not available for comment."

I paid $5, Trent Reznor paid $5000. Unfortunately, that still doesn't come to parity related to our incomes.

I paid $0, listened to one song decided I didn't like it. If I liked it I would have come back and paid for it. Because the majority of people paid nothing for it doesn't mean lost sales. I never listen to Radiohead, I only downloaded it because of all of the publicity. I always pay/download Trent Reznor's albums when he releases them.

This also doesn't take into account the people that paid X for the album and then went on to see them in concert, making them Y more (and I'm pretty sure Y>>X)

I paid $10, and then never listened to it.

I still haven't, actually.

What they always seem to miss on these pay what you want sites is a display of the average price people are paying. For example, I don't remember In Rainbows ever listing what the average price people were paying was.

If they showed a price that moved up or down with the herd, I bet you'd get more people paying that price or more, especially if you removed all the zeros from the average :)

Wouldn't removing all the zeros be a bit dishonest? An average price certainly includes all those that opted to pay $0, otherwise you don't have an accurate and honest representation of the data.

Doesn't have to be. It could just be:

Average price of those who paid at least $0.01: $5.87

Or just "Average price of those who paid: $5.87"