The Question That Makes Amazon $2.7 Billion Of Revenue(businessinsider.com)
businessinsider.com
The Question That Makes Amazon $2.7 Billion Of Revenue
http://www.businessinsider.com/the-magic-behind-amazons-27-billion-dollar-question-2009-3
6 comments
Because of the power-law distribution, on sites with substantially lower traffic than Amazon, it's very possible that the helpfulness question won't be answered enough to be useful.
How do you incentivise someone to answer the "Is this helpful?" question?
How did Amazon manage it, even? The article talks about how subtly placed the question is. Yet I know I've answered it - to thank the author of the review, or to passive-aggressively express my displeasure at the wasted ten seconds I spent reading it. I also get annoyed if people downvote reviews I spent a lot of time on because they disagree with me.
Could you assume a review was helpful because you saw that a user had read it and then purchased the item? Maybe that would help populate the rankings, albeit crudely, for sites with lower traffic.
How do you incentivise someone to answer the "Is this helpful?" question?
How did Amazon manage it, even? The article talks about how subtly placed the question is. Yet I know I've answered it - to thank the author of the review, or to passive-aggressively express my displeasure at the wasted ten seconds I spent reading it. I also get annoyed if people downvote reviews I spent a lot of time on because they disagree with me.
Could you assume a review was helpful because you saw that a user had read it and then purchased the item? Maybe that would help populate the rankings, albeit crudely, for sites with lower traffic.
Could you assume a review was helpful because you saw that a user had read it and then purchased the item?
Interesting idea. That would tend to "upvote" reviews that sell the product, pushing the negative (or even positive but ineffectual) reviews off the page. That's a good thing if what you want to do is motivate sales, but a bad thing if what you want is to promote useful reviews.
Interesting idea. That would tend to "upvote" reviews that sell the product, pushing the negative (or even positive but ineffectual) reviews off the page. That's a good thing if what you want to do is motivate sales, but a bad thing if what you want is to promote useful reviews.
Read it here to avoid the blog spam: http://www.uie.com/articles/magicbehindamazon
And yes, this is blog spam.
And yes, this is blog spam.
It is also about lowering the barrier to engaging with the product as well, right?
Facebook has done this amazingly well with first the 'comment on this story' interface, and now the like/unlike buttons.
Simple AJAXy elements like these examples seem to drive up engagement and result in positive feedback for both the receiver and the provider, resulting in driving even further interaction. Facebook probably doesn't make money but it does enhance stickiness for them.
Facebook has done this amazingly well with first the 'comment on this story' interface, and now the like/unlike buttons.
Simple AJAXy elements like these examples seem to drive up engagement and result in positive feedback for both the receiver and the provider, resulting in driving even further interaction. Facebook probably doesn't make money but it does enhance stickiness for them.
This is an interesting post. It reminded me of a post by Jonathan Schwartz from 2005 about how Sun reduced their quality programs to use single questions: http://blogs.sun.com/jonathan/entry/defining_quality
There is some insight in this article for how comment rating could work in a social bookmarking site, as for example Hacker News.
And this article is a must read if you operate a site that sells products and your site includes user-submitted product reviews.
And this article is a must read if you operate a site that sells products and your site includes user-submitted product reviews.
It's a nice discussion of the design and evolution of the helpfulness question, which I find very useful. At a deeper level it shows how a seemingly small improvement in user engagement can have dramatic impact on a site. Read it. It should trigger a re-examination of your own site.
HN thought: would the helpfulness question be better for articles than the simple upvote? Seems to me that it would?