The Economics of Variety

This is a drill-down on some issues glossed over in the short preceding FAQ on whether the Long Tail grows demand or just shifts it. One of the aims of my quantitative work with the large data sets now available from online retailers is to answer the question statistically; in the meantime, here's the academic background.

More variety does not, by itself, always encourage more demand. Indeed, as some studies have shown, people can be overwhelmed by too much choice and actually buy less, as in the 24 kinds of jam example cited by Barry Schwartz in his "Tyranny of Choice" arguments.  Amazon has found that one solution to this is to add information about the products, from something as simply as "rank by bestselling" to more detailed customer reviews and ratings, which can cut through that confusion and encourage consumers to buy with confidence. (See my previous post on why variety is not enough.)

But anecdotally, we all know of instances where increased variety and better ways to find things encourages people to consume more. I know, for instance, that Napster reawakened my interest in exploring new music, a passion that continues more easily (and legally) with Rhapsody, which has probably doubled my music spending. I also know I buy more books (and practically everything else) because of Amazon's selection. And my family certainly watches more DVD thanks to Netflix.

The army of white-earbud-wearing New Yorkers are surely listening to more music than they were before the iPod, extending the effect created by the Walkman a generation before. But are they buying more music, too? The hard numbers are, unfortunately, inconclusive. As of February, Apple had sold 10m iPods and 250m tracks on iTunes, for an average of 25 tracks sold per iPod, or less than the price of two CDs over the nearly two years  the iTunes music store had been in business.  That is not impressive.

Perhaps they're just buying CDs and ripping them. Or just downloading more, but not paying for it. Studies of the industry at large are sadly inconclusive. The RIAA says its members are shipping fewer CDs, but Soundscan says they're selling more of those they do ship (see Kensei for more on this mess).

It's not just music that's in a muddle. In books, overall book sale growth has slowed over the last decade, even as Amazon (which accounts for about 10% of the market) rose. Meanwhile, total entertainment spending has remained pretty constant between 5-6% of the average Americans' spending for decades, despite an explosive rise in entertainment options.

Indeed, although there is a general presumption that more availability leads to more sales, there is precious little statistical work that proves this is the case, especially for large numbers of products. The small-number consumer psychology work, however,  does suggest that when the choice is meaningful, more is better--you're simply more likely to find what you want, or at least something that strikes your fancy.

There are a few studies that look at examples such as the effect of increasing the number of yogurt flavors on offer by one or two, which does help sales.  One of the better-known bits of "more is better" research is Brown, Read and Summers' 2002 paper "The Lure of Choice", which conducted bank, nightclub and casino experiments to show that consumers picked options more often when they were accompanied by other choices.

Likewise, Malcolm Gladwell has highlighted the case study of the would-be spaghetti sauce competitors who learned not to out-Prego Prego in making the platonic ideal sauce, but to instead celebrate diversity--chunky, homemade, spicy--and expand the market through multiple niche sauces. In part because of this introduction of more variety, which pushed buttons consumers didn't even know they had, spaghetti sauce is now one of the top six growth categories in the dressing and sauce market.

But all this says is that some choice is better than none. What we've got in the Amazons, Netflixes, and iTunes of the world is another phenomena entirely: lots of choice turning into near-infinite choice. It is not just three hits going to 30 hits, but 300 hits going to 300,000 niches. This is a new effect, just a few years old, and I'm working with those companies to quantify it. I'm betting that the effect of massive increases in available variety coupled with good ways to find stuff is to not only shift demand more toward niches but also to grow it, following the greater-satisfaction principle. But the data will have to speak for itself.

Comments

[brian]

what I think is happening is that consumers are finding they can more closely match their preferences with substitutes to mainstream products. That is, consumer utility and hence surplus increases, but it does not change how much overall they spend on goods/content/entertainment. While some smaller niches may find an audience it may be at the expense at the margins to the heavily marketed stuff.

[Francis Hamit]

There is always a trade off in consumer choice. This is basic economics and affects policy at all levels. One of the classic examples was the trade-off chart in my old Economics textbook that demonstrated the trade off between long-range bombers and new school buildings.

There the constraint was money. Here, the constraint is time. It takes time to find the items you want and, as demonstrated by Zipf's Principle of Least Effort, most people will buy what they are looking for the first place they find it rather than look for a lesser price. This is why retail stores place all those little items next to the cash register. Availability and convenience equals more sales. For that matter so called "convenience stores" like 7/11 make most of their money from milk, bread, beer and soft drinks which they sell for far more than the price at the local supermarket. What is sold is not so much the product as the fact that it is available, right now.

Electronic distribution has two effects on this model. It widens the field of possible customers and shortens the search time. The dimunation of book sales is probably due to the rise of a more efficent aftermarket for used books through Amazon and eBay. Amazon Marketplace has over 800,000 third party providers now. That's a win-win situation. Amazon has no inventory or labor costs and makes a commission for every sale, the customers have a wider choice of available copies and the sellers are able to recycle old books at comparative advantage. What is worth 50 cents here may be worth 40 dollars someplace else.

In terms of new products, those produced in the electronic domain can be reproduced to meet all demands with almost no incremental costs beyond origination. The supply is infinite.

The cost to the consumer is less the money for the item than the time needed to consume it. Rental models like Netflix become very viable because usage and inventory are directly linked and consumers pay a very modest fee per use rather than buying and storing the same items in a collection (personal inventory).

While money cost is minimized,time cost is not and cannot be. Therefore consumer choices among a wider selection of products inevitabily means that some will sell less than if there were not so many alternatives available. Niche products very much depend on a lower cost of production and economic break-even point for their existance.

If we accept that there is only so much time and money available, the existance of wider choices represented by more niche products means that so called "mainstream" products will not sell as well.

[Alex Rowland]

The curve will most certainly flatten, but I think it is right to assume that spending on various forms of entertainment will probably not change too much from the 5-6% Chris mentioned. I think what has happened to many of these other forms of entertainment is that they have had to make way new entertainment sources (most prominently, video games) which are gradually clawing attention away from these other media. But the fact remains that while attention on entertainment will expand, revenue will not expand as appreciably. This is probably bad news for the mass market guys.

[jake]

I think you still need to ask "sales for who?" Take the music industry. While I think the long tail concept can be profitable for big retailers like Amazon, I'm not sure it allows more musicians to make a full-time living from music. So at that level, I'm not sure how sanguine I am about the long tail.

The 24 kinds of jam scenario plays out a lot in the music industry today. Before the long tail, it was probably harder for the average musician to crack through to a level where their music would be widely distributed outside of their region. But I suspect that those musicians who did crack through to this point had a reasonably good shot at sustaining a living from record sales and touring.

Today, I'm not so sure. It's so cheap to release a CD, that anyone can do it, and they do. It's also increasingly easier to make the CD available for sale nationally.

But it's much harder to actually let people know it exists, because the marketplace is so glutted. So I expect you have a lot more people selling 100-3000 CDs, which is great and it may well be that people like Amazon can make good money from this long tail of releases, but sales of 100-3000 CDs are not enough to sustain a career for most musicians.

And to the extent that digital distribution reduces the cost of production, I think this saving will be end up being spent on promotion and publicity, so it's not like that saving will go in the artist's pocket.

[jake]

I think you still need to ask "sales for who?" Take the music industry. While I think the long tail concept can be profitable for big retailers like Amazon, I'm not sure it allows more musicians to make a full-time living from music. So at that level, I'm not sure how sanguine I am about the long tail.

The 24 kinds of jam scenario plays out a lot in the music industry today. Before the long tail, it was probably harder for the average musician to crack through to a level where their music would be widely distributed outside of their region. But I suspect that those musicians who did crack through to this point had a reasonably good shot at sustaining a living from record sales and touring.

Today, I'm not so sure. It's so cheap to release a CD, that anyone can do it, and they do. It's also increasingly easier to make the CD available for sale nationally.

But it's much harder to actually let people know it exists, because the marketplace is so glutted. So I expect you have a lot more people selling 100-3000 CDs, which is great and it may well be that people like Amazon can make good money from this long tail of releases, but sales of 100-3000 CDs are not enough to sustain a career for most musicians.

And to the extent that digital distribution reduces the cost of production, I think this saving will be end up being spent on promotion and publicity, so it's not like that saving will go in the artist's pocket.

[Greg]

Your report that book sales are down really surprised me and made me try to think of reasons for this. Bookstores have always been with us, but they have gained a great deal more prominence recently, becoming huge big box warehouses with furniture and attached coffee bars, besides which Amazon, probably the web's largest retailer is still thought of by some, like myself as primarily a bookstore. My informal observations lead me to believe that people are reading more rather than less.

I really like another person's suggestion that sales being down may be a result of a better second hand market, the first thought that came to my mind is that there are many personal websites and forums which effectively displace the need for books of some kinds.

Where previously they sold books of household hints and handy tips they have websites where millions of tips are solicited from readers around the world. An infinite supply of recipies is available online and instruction and tutorials on how to use software, like Adobe Photoshop.

Since the expansion of the internet a lot the reasons we used to buy and keep nonfiction books we now turn to websites maintained by hobbyists or supported by advertising.

[christmas gift ideas]

Great review and sharing.I have read this post and just
wanted to say this theory is obsessed with the media.The content are not clear in it.I would like to request please share something about the borrow of phrase in the medium.Thanks.