Talk:The Long Tail
From Wikipedia, the free encyclopedia
The image of the long tail (grey background w/ red and yellow plot) is also found here: http://longtail.typepad.com. I'm not sure who copied the picture from who, but this could be some sort of copyright violation. 68.113.132.119
- The image was uploaded with the permission of the creator. Stbalbach 03:56, 12 Mar 2005 (UTC)
Gracious.. is there any way to make the opening paragraph 1) shorter 2) friendly to the general reader 3) encourage and invite the reader to want to read more? The amount of detail and technical talk in the opening paragraph should be very limited .. its purpose is to provide an overview of whats contained in the article body, a summary in simple language, a broad context, inviting the reader who wants to learn more to read further... like peeling an onion.. it really is not a complicated or difficult concept and should not be made so unless the reader wants that level of detail.. --Stbalbach
Certainly a problem. The key trouble in the current opening is the word "distribution" which will scare off anybody who's even slightly math phobic. The easy out is to let the entry by about just the one example that Chris Anderson is hot on the trail of; i.e. the long tail of products. That is probably a bad idea because there are so many domains where recognizing the long tail helps to illuminate the discussion. For example "the poor" is the name given the long tail in of the wealth distribution. Possibly we could drag the second paragraph of examples up and introduce them not as examples but as exemplars of a curious phenomenon "A curious pattern has been noticed in diverse systems..." that is then given a more formal name in the second paragraph... "Statistics geeks have names for these things..."
Of course the Anderson meaning for Long Tail is currently dominate in the rest of the article. But the term didn't really have that business buzz meaning until he wrote the article. Prior to that it's most common usage was in wealth, blogs, city size, and vocabularies. Probably in that order.
This is a general problem; how to get terms like this that live on the border between popular science (or even fads) and formal science. The articles on things like power-law aren't particularly user friendly, which is a bummer.
- Bhyde
-
- Thanks for the info, didn't realize "long tail" was in common use before the Wired article, Google search confirms multiple sources (date restricted to pre-Wired article publication date) [1] .. so we need to restructure the article to better reflect the multiple uses, meanings and implications of the term, depending on the domain of use. It can still be user friendly in the opening (the concept is not difficult), but there should be a seperate section for Chris, and a section for the scientific explanation, etc.. --Stbalbach 07:05, 3 Feb 2005 (UTC)
Chris Anderson here. I'm pretty sure I coined the term, at least in the current proper-noun sense. Of course there had been discussion of the "tail" of curves before, and the observation that some distributions have long tails is not new, either. But "The Long Tail" as a stand-alone description of the universe of niches is my own construction, begining in a series of speeches in early 2004 and culminating with the publication of the Wired article. Although there do appear to have been some uses of the term in an economics sense before the piece, I think it's fair to say that its current popular meaning derives from my article and this entry should reflect that. --Zlite 01:12, 5 Feb 2005 (UTC)
Ben now ... From my corner of the world it certainly appears that Clay Shirky gets a lot of cred for dragging the term "long tail" into the discussion of the "norati." His very widely discussed essay on power-laws across the blogs uses the term in 2003. It then get's widely used as a way to think about what is sometimes called "citizen media."
But it's not who popularized it that concerns me. What concerns me is that Chris's meaning is tied to taking a sample over a space of products. I find leaving out the blogs very troubling; since the structure of "citizen media" is driven by this effect. It's key to how Wikipedia and the like draw on huge swaths of contributors. I use the term at least twice on my blog in 2003 in discussions about how contributors to open projects, creative markets, and mailing lists are distributed and to make clear that the value contributed from that tail is key. For example I use it to talk about how when font foundries were disrupted by TrueType the long tail of creative designers was key. For example I use to talk about the long tail of contributors to open source projects quite a few times in various ways. For example Microsoft's developer network is all about harvesting the creative energies of a long tail of developer labor.
Focusing the term in on the business model of Amazon, eBay, and Netflix is a boo boo.
-- Bhyde
A case could be made that "The Long Tail" is the name of a magazine article, a blog site, a possible forthcoming book, and a popular concept and thus deserves its own Wikipedia article. It's an article that goes beyond a description of a statistical phenomenon. As well, an aricle called "long tail" (without the "the" and in lower case) should describe the statistical phenomenon and historical use of the term beyond what Chris describes, including Clay Shirky (if he qualifies as expert?). Just some thoughts.
--Stbalbach 23:02, 6 Feb 2005 (UTC)
I think I've made the case that "The Long Tail" as a popular neologism predates Chris's work to focus attention on the particular domain of shelf space. Since they are key to understanding world poverty, the value globalization, the function of open systems like wikipedia and open source, even culture I certainly don't want to rain on Chris's parade. I am a strong supporter of popularizing and spreading understanding of how wide spread the syndromes are around the long tail.
But, I think it's a mistake for the ...pedia to adopt Chris's current definition as definitive. It isn't durable. It's too narrow and it's too tied to the buzz he's spinning up. I love the buzz, I just don't think it serves the purposes of the ..pedia to become tied to it.
It appears we have at least four claimants to the term. The Blog space claim (ala Clay), the Shelf space claim (ala Chris), the statistical/mathematics claim, and finally the claimant as a generator of value in social systems. There is probably a five claimant which is meerly an enumeration of striking exemplars. Each of these domains the definitive "the" get's used to mean different things.
That pattern is not unusual with terms, witness the # of entries on most any work in the OED; but it's particularly severe with neologisms.
-- Bhyde
For my part (clay shirky here) when I used the term in Weblogs, and Inequality, I didn't think of it as a coinage at all -- linear distributions have heads and tails and the weblog tail is long and flat. As Ben points out, the phrase has been widely used long before now (as we'd expect of two common words in Englich that describe a coherent and widespread phenomena.)
But that kind of originality isn't what's in question, anymore than whether Malcolm Gladwell coined "tipping point" -- he plainly didn't, but he associated that phrase in people's minds with a whole class of phenomena, and did it so successfully that its now hard to have a conversation about at least some of those phenomena without using that phrase (which is frustrating in some ways, as there is never a "point" where anything happens on logistic curves.)
So, I think, with the case of the long tail and The Long Tail. Chris and I and lots of pther people use the phrase to describe a particular kind of distribution, but Chris has taken it in the direction of Tipping Point, a phrase that conjures up a whole complex of related issues, particularly issues of the business aspects of media and culture, that I didn't. So from my pov, Chris should get credit for originality, not of the phrase but of its current application and vividness. This colloquialization will sometimes be problematic for people trying to have more precise conversations, but language is like that, and it seems likely to me that the Long Tail will enter the general lexicon through Chris's work. -clay
Ben again... Sigh. Chris is not using the phrase to describe a particular kind of distribution. From the about section on thelongtail.com: "The Long Tail is the yellow part of the sales chart at left, which shows a standard demand curve that could apply to any industry, from entertainment to services. The vertical axis is sales, the horizontal is products. The red part of the curve is the "hits", which have dominated our commercial decisions to date. The yellow part is the non-hits, or niches, which I argue in the article will prove equally important in the future now that technology has provided efficient ways to give consumers access to them." Chris is using it to describe a syndrome in sales and marketing, a class of business plans. Is a vivid article, a domain name, and a half million dollar book deal the threshold of owning a neologism? I guess so. The harm as I see it is that we will fail to draw upon the understanding that arises from seeing that this pattern appears in ecology, sociology, geology, etc. in effect privatizing the term for that branch of economics occupied by venture capitalists. Certainly this is analogous to the tipping point example. But in that case the term wasn't being dragged off and stripped of a wealth of it's meanings; rather it was being illuminated for the general public. I'm excited that Chris is illuminating this for the public; I'm concerned that one very narrow business planning meaning is where his light has fallen. - Bhyde
Interesting discussion on a number of levels. Ben, I think you may be fighting a losing battle here. Human language is probably the most organic thing in the universe that's not actually alive. As such, nobody gets to decide which neologisms or, as the case may be, solecisms, ought to mean what. This is less a 'neener-neener' than a reality check. As authors of this entry, our responsibility is to supply readers with the definition of the term they are looking up -- *not* the ideal definition (i.e. one that we decide will be more "durable" or inclusive or logical.) The good news is, if the narrow definition becomes a part of the lexicon, it will soon expand of its own accord to fit all the phenomena people see as analagous or equivalent. The various histories of the term are clearly of interest, but the important thing here is to give people a simple and elegant definition of the specific contextualized phrase they're looking up. - HarpooneerX
Contents |
[edit] The Long Tail, Long Tail, or Long tail?
I'm pretty sure this article should move, as the initial article doesn't pass the "capitalised in running text" test, nor does it seem to be specifically the name of a work, but I'm in two minds as to where. Is the "proper noun" sense really primary? I'd be inclined to believe it just had more media coverage, but I'm willing to be convinced otherwise. Alai 23:43, 27 May 2005 (UTC)
- A case could be made for all three. The article is a composite of different meanings. The proper noun usage is the most well know in the wider public sphere. Stbalbach 01:17, 28 May 2005 (UTC)
[edit] sorting by popularity
When you sort items by popularity (or size, or ...), the most popular ones (or biggest, or ...) always come first, followed by the less popular ones.
Is that all that "the long tail" is about ?
Or is it about the following counter-intuitive fact: ? Sometimes (not always), the small rocks outweigh all the big rocks put together, even though any one big rock outweighs any one small rock.
For example, How Long Is the Coast of Britain? Statistical Self-Similarity and Fractional Dimension.
The article mentions Lots of energy was released by the earthquake of December 26 2004, but there are tiny earthquakes all the time; most earthquakes are part of the long tail.
Yes, but can we go even further and claim that all the tiny earthquakes, put together, release more energy than the big earthquake? Is that true ?
--DavidCary 19:24, 11 Jun 2005 (UTC)
This statement in the article expresses a myth rather than a fact about the Long Tail: "In many cases the infrequent or low-amplitude events—the long tail, represented here by the yellow portion of the graph—can cumulatively outnumber or outweigh the initial portion of the graph, such that in aggregate they comprise the majority." It is almost never true in any real world distribution. In these situations an 80-20 or Pareto distribution applies: 80% of whatever comes from 20% of the users. Likewise, 80% of the users cumulatively account for only 20% of whatever it is you are measuring. The Long Tail represents opportunity in some sense, but it is a total misrepresentation to suggest that the Long Tail represents more opportunity than going for the masses.
--underalms 03:15, 23 Feb 2006 (UTC)
The more I think about it, the more I am convinced that this long tail effect is nothing but the next "paradigm shift" and has no meaning independent of the book being written. We seem to be suggesting that there are situation, as underalms noted, where the tail has more probability mass than the area around the center. But that's pretty much nonsense in most satistical distributions created from real world data. In the business sense, the long tail argument boils down to nothing more than economics of scale. If your example companies were truly boutique retailors who refused to serve the center of the market and were successful, you might have a point. But Amazon or Netflix aggresively pursues the entire market, the only reason they can offer more selection is because of their cost structure. Netflix doesn't offer you the chance to rent "Ernest Goes To Camp" because the collective demand for all the Ernest movies excedes that for "Batman Begins" but because they store many times more titles due to their lower cost and that allows them to offer titles that have very low demand. -BennyAbelard Aug 25 2006.
[edit] Merge tag
That should not be done. "The Long Tail" as a proper noun with capital letters as first popularized by Chris Anderson goes way beyond any statistical definition. If Chris had called it "The Long Toenail" would you still suggest merger? The fact they have the same name doesnt mean they should be in the same article, they are different concepts. Stbalbach 15:41, 29 September 2005 (UTC)
Please don't merge the two, they're completely different concepts. People suggesting merging should probably understand what the two means before suggesting something this absurd!
[edit] Fillums
Would the post-mortal success of films such as "John Carpenter's The Thing", "It's a Wonderful Life", "Night of the Living Dead" and so forth be a good example of a "long tail"? "Donnie Darko", there's another one. I suppose it would only count in a single form of market, i.e. video distribution cannot be a "long tail" for cinematic exhibition, because they are different markets entirely. Perhaps a better example would be films which open poorly in several cinemas but then play continual packed houses in one or two cinemas, such as "The Rocky Horror Picture Show" or... "The Exorcist", in Britain, when it was banned on video. -Ashley Pomeroy 14:05, 22 October 2005 (UTC)
[edit] English word frequency
The numbers that appear in "Such distributions are surprisingly common. In standard English, the word the..." does not look correct. The link that is given does not contain such information (frequencies) and many other links contain completely different numbers. Example: http://www.duboislc.org/EducationWatch/First100Words.html
MeirM 22:36, 14 January 2006 (UTC)
[edit] This Article May Not Fit in the Long Tail
Ironically, this was one of the top results for articles when I did the google search "From wikipedia, the free encyclpedia" in quotes with "site:wikipedia.org" -- which turns up english articles with high ranking in the google page indexes. So the quote from this article "while this page might be on the far right in the yellow," probably doesn't apply. Or does it? —The preceding unsigned comment was added by 65.19.220.226 (talk • contribs) .
- Interesting, never heard of that Google trick to find a way to rank Wikipedia articles. Yeah, ironically the book or the meme in general is not in the long tail either. -- Stbalbach 14:30, 20 October 2006 (UTC)
[edit] Pizza, Television and The Long Tail
What type of Pizza is usually on sale? Pepperoni. Is pepperoni usually on sale because it's the most popular, or is it the most popular because it's usually on sale? What about all the other pizza toppings? Are they collectively a massive Long Tail in sales?
Television networks like to cater to a very narrow range of ages and what types of shows the TV ratings tell them that narrow range likes. But they're missing the big market in the highly differentiated Long Tail.
One only needs to compare the postings on the official FOX board for Firefly VS their board for The X Files. The number of posts on the Firefly board nearly equalled the total number of posts on The X Files board (which had been available for several years) before Firefly was cancelled.
Some data mining would've provided to FOX some very good statistical information on exactly who was watching Firefly. Instead, FOX chose to use the Nielsen Media Research data, which at that time (except during 'sweeps') polled only between 10,000 and 15,000 housholds each week, and did not include data from other sources such as TiVo (I've read that NMR recently cut a deal with TiVo for their data) which reported that Firefly was at times the #1 show put on Season Pass to record every episode.
Ignoring the data at the ends of a bell curve graph can be ignoring the largest portion of the market for your company. Focussing only on the big lump in the middle is easy, but there's a lot of 'fine gold' to be found in The Long Tail!
Killing a product that doesn't immediately show a large interest from the narrow range is a dumb move. Instead, figure out how to shift the focus to the people who are buying it.
Even worse is killing a product that shows a large interest, but not in the market segment you targeted. An example is the Nickelodeon animated TV series "Invader Zim". The target was pre-teen children but the largest viewer group was older teens and adults. (The exact market other networks love to have such response from.) Instead of taking advantage of that by moving the show to a timeslot even better for the audience it attracted, Nickelodeon cancelled it. Chasing the tail then cutting it off when you find out that the tail doesn't care but the body likes your product and wants more is just as dumb as ignoring the tail completely. —The preceding unsigned comment was added by 67.136.145.205 (talk • contribs) .
- Pizza sounds like a good field study. -- Stbalbach 15:35, 26 October 2006 (UTC)
[edit] Save the Long Tail
Is there some way to incorporate into this article something about the proposed changes to the internet by big media companies who want to charge extra for websites to use "their pipes"? The whole thing regarding net neutrality...
Cuz the long tail will be almost totally curtailed if net neutrality is not maintained.
What do people think? I know some nit-pickers are going to say this is opinion, but in reality I think it is fact that the long tail will not be possible since the cost of inventory, the cost of having a website that people can actually view in a reasonable period of time will be higher than what would allow for the long tail.
209.162.21.173 22:30, 29 October 2006 (UTC)
oops, forgot to sign in. this was my post Misterman8 22:34, 29 October 2006 (UTC)
- If you can source it and attribute who talked about it. It's probably not the right article to go into much detail, network neutrality would be the better place to expand on it. -- Stbalbach 15:15, 30 October 2006 (UTC)
