Amazon and Pepsi: Thanks for the memories

When others were readying to ring in the new year, I spent more than an hour scrambling to use up my last 94 points from the Pepsi Stuff promotion. As promised, the website and Amazon’s page disappeared on Jan. 1, even if the Facebook page remains live.

I think the promotion was well designed and well executed. Announced in January and launched with a Super Bowl ad, the promotion began Feb. 1 and ran for 11 months.

This was Pepsi’s second effort to seed digital downloads, following its earlier cooperation with Apple’s iTunes Music Store. With its 2003 press release and a 2004 Super Bowl ad, Apple and Pepsi announced plans to give away 100 million songs, but only about 5 million were redeemed. A second promotion (twice as big) ran in 2005, although both the first and second contest allowed cheating.

Both the original Pepsi promotion and the 2008 successor were effective with me. In the most recent case, I signed up with Amazon, installed its downloading application, and bookmarked its webpage for weekly “MP3 Deals”.

Mainly through my wife’s (two case/week) Diet Pepsi habit, at one point per sixpack I accumulated 259 points: 180 for songs (36 songs), 75 for a t-shirt, and 4 that went unused. I downloaded 36 songs with Pepsi Points — 18 in the final four hours of the promotion. 21 songs were classical albums (one song at a time), most of the rest were singles from groups where I didn’t want the albums.

But the promotion was also effective at shifting my attention during this period from iTunes to Amazon. In some cases, the cost per song on the MP3 Deals was too good to pass up: I bought five albums (with 9-21 songs each, most at $5): two classic rock albums, one classical album and a Xmas album each for my wife and daughter. I bought five singles, three of those Xmas songs.

I originally thought the UI was clunky compared to iTunes, and there were serious problems due to the web page metaphor. On multiple occasions, I unintentionally interrupted the download (but not the purchase) by backing up in the browser, leaving me with no song; tech support had to manually make the song available. This is something that would be easily fixed if every purchase opened a new page.

Because Amazon requires one click purchase for songs — something I otherwise disable on my Amazon account — I also accidentally bought one song I had no intention of buying because I clicked on the wrong part of the UI. (Given we’re in America, this seems like a lawsuit waiting to happen when someone with poor vision or motor skills claims fraud due Amazon’s policy.)

On the other hand, I thought the Amazon merchandising was much more clever and aggressive than Apple’s staid and dated site. Apple offers one free song a week, which (I expect) will always be something oriented at its teen demographic. Amazon offers dozens of free songs, which caused me to linger longer at the site; the $5 albums certainly caused me to keep coming back every week, particularly since one was usually a classic rock album aimed at my demographic (such as “Dark Side of the Moon,” which I once owned in college but never acquired as a CD).

Both sites have restrictions on downloads, many of them obviously imposed by the labels. The premium classical brand, Deutsche Grammophon (one of four classical music subsidiaries of Universal Music Group), marks at least one song of every piece (e.g. symphony) as “album only,” forcing customers to buy the entire album. On the other hand, Amazon’s variable song pricing makes sense for longer cuts (such as a 15-minute movement from a classical piece) than the standard 3-5 minute single on a pop album.

Amazon also has DRM-free, while Apple’s efforts to shift to DRM-free have stalled after only EMI decided to cooperate. The iTunes DRM system works reasonably well, but every so often is intrusive to the point of being irritating.

Apple is still the market leader in music downloads, with five billion songs as of last June and perhaps an official announcement of a new total (seven? eight? billion) next week at Macworld. As with the Kindle, Amazon is not saying how many songs it has provided, only the number of songs available.

Overall, I think Amazon managed this promotion very effectively, using it to reach second place in the (U.S.) download market. Still, it’s not clear whether I’ll be back to the Amazon store now that I no longer have Pepsi Points to spend.

Digital music files are a commodity for me: it doesn’t matter whether I download them from Amazon, from iTunes, or buy a used CD and convert it: the main driver is price. The CD is usually cheapest if I want an album, while the downloads are cheapest if I only want one song (assuming the publisher allows it to be sold ala carte). It’s easier to compare prices between Amazon and iTunes (or between Amazon and its used CD sellers) than it is to go to a physical Wal-Mart and Target, and anything that makes price-shopping easier fuels commoditization.

The record labels have done everything they can to enable Amazon’s entry to fuel competition to Apple and reduce its channel power. To the UCLA and USC MBAs running the record labels, this must have seemed like a no-brainer application of the “supplier power” concept of Porter’s Five Forces.

However, lacking a quality or easy of use advantage — or any sort of differentiation other than DRM-free and integration with the Amazon catalog/buying system — Amazon has been aggressively using price to gain market share: free songs, bargain albums, songs at 89¢ or 79¢ each. This further encourages shoppers to price shop, reducing the perceived value of digital downloads and increasing commoditization.

Between pirated content and distribution alternatives, the music industry is further along the commodization path than is its video counterparts. I can’t see a scenario (even in the short run) where the labels can push through a price increase (as does the movie industry) to Amazon and iTunes, and with commoditization I’d expect music publishers’ average per unit revenue in real dollars (not to mention unit sales) to be lower a decade from now than it is today.