The internet was supposed to bring vast choice for customers, access to obscure and forgotten products - and a fortune for sellers who focused on niche markets.

But a study of digital music sales has posed the first big challenge to this “long tail” theory: more than 10 million of the 13 million tracks available on the internet failed to find a single buyer last year.

The idea that niche markets were the key to the future for internet sellers was described as one of the most important economic models of the 21st century when it was spelt out by Chris Anderson in his book The Long Tail in 2006. He used data from an American online music retailer to predict that the internet economy would shift from a relatively small number of “hits” - mainstream products - at the head of the demand curve toward a “huge number of niches in the tail”.

However, a new study by Will Page, chief economist of the MCPS-PRS Alliance, the not-for-profit royalty collection society, suggests that the niche market is not an untapped goldmine and that online sales success still relies on big hits. They found that, for the online singles market, 80 per cent of all revenue came from around 52,000 tracks. For albums, the figures were even more stark. Of the 1.23 million available, only 173,000 were ever bought, meaning 85 per cent did not sell a single copy all year.

Mr Anderson told The Times yesterday that he accepted that Mr Page and his co-researcher, Andrew Bud, the head of mobile software company mBlox, had found a dataset in which the “long tail” principle did not apply. But he said further conclusions could not be drawn until the data and its sources were published.

Mr Page and Mr Bud believe, however, that their findings seriously undermine Mr Anderson’s thesis, which came with subtitles such as: How endless choice is creating unlimited demand and Why the Future of Business is Selling Less of More.

“I think people believed in a fat, fertile long tail because they wanted it to be true,” said Mr Bud. “The statistical theories used to justify that theory were intelligent and plausible. But they turned out to be wrong. The data tells a quite different story. For the first time, we know what the true demand for digital music looks like.”

Mr Page, who carried out the economic modelling for Radiohead’s In Rainbows album, which was released free on the internet, said: “The relative size of the dormant ‘zero sellers’ tail was truly jaw-dropping. Rather than continue to believe the selective claims of ‘here’s another great example of the long tail at work’, we wanted to find out how longtail markets should be analysed, plotted and interpreted.”

However, Mr Anderson named by Time magazine as one of the world’s 100 most influential people - told The Times: “There is a reason why the ‘long tail’ has become a fixture in the technology world over the past five years - it fits countless phenomena we see every day.

“I respect what Will’s done and have no doubt that he has indeed found a dataset where it doesn’t work, but I’m not sure you can conclude much, if anything, beyond that. If he’s trying to undermine the entire Long Tail Theory, he’ll have to provide a lot more evidence. I welcome the debate, but until Will’s prepared to publish data and sources we don’t have much to talk about.”

Mr Page and Mr Bud found that, rather than following Mr Anderson’s predictions, online music sales followed instead a sales distribution laid down by Robert Goodell Brown, an American economist, in 1956.

Mr Brown, who was an academic at Yale, outlined the theory in Statistical Forecasting For Inventory Control, a landmark tract on inventory control that focused on the sales of industrial items such as rivets and widgets.

Mr Page said: “There is an eerie similarity between a digital and high-street retailer in terms of what constitutes an efficient inventory and the shape of their respective demand curves. I think there’s something more going on there: a case of new schools meets old rules.”

Mr Page and Mr Bud are working on a book of their findings and hope to stage a debate with Mr Anderson in Brighton next May.

The long and short of it

—Chris Anderson’s The Long Tail challenged the “80/20 rule” widely accepted in retailing. This suggests that selling the most popular 20 per cent of products is the way to make a profit as they will account for 80 per cent of sales

—Anderson’s analysis of online music sales suggested that, thanks to the cheapness, simplicity and global accessibility of searching for products online, retailers could make money from more obscure products because they would always find an audience

—An employee of Amazon, a company seen as an example of the Long Tail Theory in practice, once said: “We sold more books today that didn’t sell at all yesterday than we sold today of all the books that did sell yesterday”

—Eric Schmidt, the CEO of Google, described The Long Tail as “brilliant and timely”. Malcolm Gladwell, the writer and sociologist, labelled it a “Truly Big Idea”

—Critics suggest the book is more a projection of an idealised market place than a model of a real one

Explore Music

• CD Reviews

• Live Reviews

• Classical

Editor's choice

• The 100 Best Films of the Decade
• 70 facts you didn't know about Marvel
• The 50 Biggest Movies Of 2010