Digital has not marked the death of the album, but labels and artists should re-think how they sell music, according to a new study. Working Knowledge has a Q&A with Professor Anita Elberse of Harvard Business School that discusses her latest working paper, "Bye Bye Bundles: The Unbundling of Music in Digital Channels." The name may sound familiar. Her article "Should You Invest In The Long Tail?" challenged the theories of the book "The Long Tail" and received a good amount of attention.

By analyzing the data of a random sample of SoundScan data of over 200 artists from January 2005 to April 2007, Elberse sought to find out what happens when people start buying digital music. The answer will come to no surprise to anyone in the music business: consumers started buying individual tracks instead of entire albums. Informally called cherry picking, this tendency of consumers to pick a song or two instead of the entire album is an erosive force to revenues that has not subsided.

Elberse's paper is deeper than this obvious nod to format substitution. Among her findings for the sample of artists she studied:

-- A drop of about one-third of the total weekly sales of an album and its songs can be tied to consumers switching from physical to digital formats.

-- For every 1% increase in the downloading rate, there is a 6% decrease in album sales and a 9% increase in single track sales per bundle of an artist's music. That means that as more people start buying digital music, they spend less money by buying fewer albums and more tracks.

-- Albums that have one or two standout tracks are more susceptible to cherry picking than albums with songs of more even quality. In other words, two great songs plus eight stinkers will be an underperforming bundle of music.

-- Albums with more songs are not at an advantage over songs with fewer songs. An album with 18 songs has a lower per-track price than an album with ten songs, but consumers were just as likely to cherry pick from an 18-track album as a ten-song album.

The paper offers many important takeaways for labels, managers and artists and should prompt them to re-think how they approach the packaging of recorded music.

-- Adding songs to an album will not reduce cherry picking. It may seem like you're giving consumers more bang for their bucks, but Elberse's research shows consumers are no less likely to upgrade to an album with 14 songs, for example, than upgrade to an album with ten songs. To get consumers to buy an entire album instead of a few tracks, other approaches (such as track prices relative to album prices) may work better.

-- One or two songs and a bunch of filler will put an album at a disadvantage. Once the album was unbundled at digital stores, consumers were able to avoid the lower quality songs altogether. In the digital age, an album with songs of similar quality should fare better than an album with larger deviations in quality from one song to another.

-- Because song quality tends to vary (rare is the album that is great from start to finish) consider different types of bundles, or removing the consumers ability to unbundle an album.

-- Since most albums are promoted via singles or, at the very least, emphasis tracks, consumers are going to be more attracted to some songs more than others. That means the consumer has to decide whether or not to spend considerably more money on additional, less familiar songs when opting to purchase an album over a track. There's a big difference between $1.29 for a track and $9.99 for an album, but less difference between $1.29 and $6.99. What are those extra tracks worth to the consumer? Consider these issues when creating and pricing bundles of recorded music.

Elberse summarizes much of her findings here:

I think labels should rethink the essence of a bundle. An album with around 12 songs may be a fine format for some artists, but why would it necessarily fit the majority of musicians? Digital channels give labels great flexibility to try alternative formats. My results show that giving preference to quality over quantity and designing smaller, more consistent bundles may be beneficial.

Follow Billboard senior analyst Glenn Peoples on Twitter at