Is Justin Timberlake a Product of Cumulative Advantage?(nytimes.com)
nytimes.com
Is Justin Timberlake a Product of Cumulative Advantage?
http://www.nytimes.com/2007/04/15/magazine/15wwlnidealab.t.html?ex=1334203200&en=79be2f770fc76c6d&ei=5124&partner=permalink&exprod=permalink&pagewanted=all
10 comments
I blogged this last night about how it pertains to what we're doing:
http://www.zaid360.com/?p=143
Cool feature of NYT online I just discovered: If you double click a word, it'll define it for you in a pop up window.
And closer to home: Is Google?
Is Java?
Lisp?
Granted, there are genuine "critical mass" effects for programming languages, but the same can be said to an extent of pop culture. If one of the reasons people listen to music is to have something to discuss with their friends, then of course, they go toward the already established.
While the details are interesting, "cumulative advantage" looks like a new buzzword for an old concept. As if it was heretofore believed that people thought independently of one another...
Lisp?
Granted, there are genuine "critical mass" effects for programming languages, but the same can be said to an extent of pop culture. If one of the reasons people listen to music is to have something to discuss with their friends, then of course, they go toward the already established.
While the details are interesting, "cumulative advantage" looks like a new buzzword for an old concept. As if it was heretofore believed that people thought independently of one another...
How so?
If you want close to home, think news.yc or other social news sites when it comes to cumulative advantage.
If you want close to home, think news.yc or other social news sites when it comes to cumulative advantage.
Your point is that news.yc and other social news sites are actually more useful when more people use them, right? I agree with that. But IMHO, that's missing the point of the article.
Music doesn't objectively get better just because more people listen to it; and yet, the authors' research appears to show that people's opinions of music is hugely influenced by what other people like. Recall their main result: They had one group of users who they didn't show any information about how often a song was downloaded, and then eight independent groups who were shown the download count of each song inside that group. "The song 'Lockdown,' by 52metro, for example, ranked 26th out of 48 in [the group not shown download counts]; yet it was the No. 1 song in one social-influence world, and 40th in another. Overall, a song in the Top 5 in [the group not shown counts] had only a 50 percent chance of finishing in the Top 5 of success." Their conclusion is that the reason why predicting the success of media products is shaky is that what's a success is hugely influenced by an unpredictable feedback process.
Thus, my question: Can we assume that Google's success is due in a large part not to anything that Google did, but to the (random) feedback process of people liking Google because they see other people liking Google?
Of course, one could argue that this is simply the truism that one needs luck as well as talent to succeed. But I come away from reading the article with the feeling that I've learned something that goes deeper than that, and I'm trying to figure out how to apply it to the tech world.
Music doesn't objectively get better just because more people listen to it; and yet, the authors' research appears to show that people's opinions of music is hugely influenced by what other people like. Recall their main result: They had one group of users who they didn't show any information about how often a song was downloaded, and then eight independent groups who were shown the download count of each song inside that group. "The song 'Lockdown,' by 52metro, for example, ranked 26th out of 48 in [the group not shown download counts]; yet it was the No. 1 song in one social-influence world, and 40th in another. Overall, a song in the Top 5 in [the group not shown counts] had only a 50 percent chance of finishing in the Top 5 of success." Their conclusion is that the reason why predicting the success of media products is shaky is that what's a success is hugely influenced by an unpredictable feedback process.
Thus, my question: Can we assume that Google's success is due in a large part not to anything that Google did, but to the (random) feedback process of people liking Google because they see other people liking Google?
Of course, one could argue that this is simply the truism that one needs luck as well as talent to succeed. But I come away from reading the article with the feeling that I've learned something that goes deeper than that, and I'm trying to figure out how to apply it to the tech world.
I think what Daniel is saying is that on social news sites, popular posts tend be popular because they're popular.
I don't think we can attribute Google's success to cumulative effects:
I suspect that you can see cumulative advantage effects only when objective quality is difficult or impossible to measure. Taste in music is extremely subjective - it's very difficult to measure the quality of music independent of its popularity.
On the other hand, the objective quality of search results can be measured by independent testing and feedback - it's not perfect - but you can get a much better measurement of quality for search than you can for music.
I don't think we can attribute Google's success to cumulative effects:
I suspect that you can see cumulative advantage effects only when objective quality is difficult or impossible to measure. Taste in music is extremely subjective - it's very difficult to measure the quality of music independent of its popularity.
On the other hand, the objective quality of search results can be measured by independent testing and feedback - it's not perfect - but you can get a much better measurement of quality for search than you can for music.
Search is a business with notoriously small lock-in and network effects. Eric Schmidt comments on this all the time, it's a significant risk factor for Google.
Brand loyalty is a cumulative advantage, but it's pretty fickle. Customers will stick with a trusted brand as long as they continue to trust it, but all it takes is one or two screwups or a competitor coming out with a noticeably better product, and your brand evaporates.
I first used Google in 1999, and it was much better than the existing search engines. Not a little better, not even just noticeably better, but orders of magnitude better. I could immediately find results that I'd previously have to dig through 10 pages of links for.
They may not have that advantage now, but I don't care. Google is good enough. Until someone comes out with a search engine that's noticeably better, I'm not going to switch. I've already got my Firefox toolbar setup, and my habits - I see no reason to reevaluate them until there's clear evidence that I can do better.
Google almost pulled off the same thing with GMail, but Yahoo and Microsoft stopped them. When GMail came out, it was orders of magnitude better than the existing webmail solutions. It gave you 1GB of space instead of 4MB, it had Google search integrated, and it had a clean AJAX UI. Like many early adopters, I switched as soon as I could get my hands on an invite (you only got 1 or 2 back then), and have since made GMail my default e-mail client. But it didn't completely penetrate the mass market, because Y!Mail and Hotmail immediately upped their storage space to 400+ MB, and Yahoo came out with a new UI. IMHO, GMail is still better, but it's no longer an order of magnitude better, so people who are just getting GMail invites now often choose to stick with their existing e-mail.
Brand loyalty is a cumulative advantage, but it's pretty fickle. Customers will stick with a trusted brand as long as they continue to trust it, but all it takes is one or two screwups or a competitor coming out with a noticeably better product, and your brand evaporates.
I first used Google in 1999, and it was much better than the existing search engines. Not a little better, not even just noticeably better, but orders of magnitude better. I could immediately find results that I'd previously have to dig through 10 pages of links for.
They may not have that advantage now, but I don't care. Google is good enough. Until someone comes out with a search engine that's noticeably better, I'm not going to switch. I've already got my Firefox toolbar setup, and my habits - I see no reason to reevaluate them until there's clear evidence that I can do better.
Google almost pulled off the same thing with GMail, but Yahoo and Microsoft stopped them. When GMail came out, it was orders of magnitude better than the existing webmail solutions. It gave you 1GB of space instead of 4MB, it had Google search integrated, and it had a clean AJAX UI. Like many early adopters, I switched as soon as I could get my hands on an invite (you only got 1 or 2 back then), and have since made GMail my default e-mail client. But it didn't completely penetrate the mass market, because Y!Mail and Hotmail immediately upped their storage space to 400+ MB, and Yahoo came out with a new UI. IMHO, GMail is still better, but it's no longer an order of magnitude better, so people who are just getting GMail invites now often choose to stick with their existing e-mail.
One word: MySpace.
It may be that a large part of the success of Facebook comes from the way they were able to use student bodies as small, self-contained networks, and then build from there.
It reminds me of building a fire with charcoals. If you keep them together you can keep them burning. If you can get a nice group going you can add to it bit by bit and build as big as you want. If you disperse them they die.
Our world has a million little self-contained networks. If you try to win some over perhaps some will succeed to the point that the network works for you. Then maybe you can add others.