Profiles in Popularity I--A Statistical Analysis of The Twenty Most Popular Reviewers
Oct 18 '00 (Updated Nov 12 '00)
Note: When I first posted this, I had no idea how very close the Presidential race was going to be, or that the winner of the popular vote may well not be our next President!
The U.S. Presidential race, among other things, is a popularity contest. Though most people want to be popular (if they will admit it), by its very nature popularity is exclusive. Only one person will win the Presidency, and only a handful of Epinions members will be among the most popular. Yet everyone starts at the same place (though at different points in time) with a new Epinions membership and a first review to write. Some people persevere and become popular, while others do not.
What is the essence of popularity at Epinions? In this review, I analyze and comment on reviewer popularity by examining the public profiles of the 20 most popular reviewers at Epinions. Who are these people--the Most Popular Reviewers? According to the Community Center, the 20 most popular reviewers ranked by total visits are: BrianKoller, ptiemann, poseidon, drlolipop, forkids, Andrew_Hicks, vara, 1GQ, hwz1, Profilewriter, kornell911, charles, lap0530 (blushing), netKat, Bonies7, mattjoe, poneyby, chrisbickel, pweisenberg, and erin5oaks.
A Personal Case Study
Since joining Epinions on January 8, 2000, I have written 262 reviews--this makes my 263rd. That averages out to almost one review per day, which as you will see is about average for a Most Popular Reviewer. I had the great good fortune to be selected as a Computer Expert (now termed Advisor) in March 2000, and found to my pleasant astonishment that my reviews were being read by both Epinions members and visitors.
To my complete amazement, my name recently began to show up on the Most Popular Reviewers page in the Community Center, and as of the most recent posting there--if Epinions’ numbers are to be trusted--my computer reviews have now been read more than anyone else’s. This is an incredible honor to me, and the popularity I have experienced is tremendously gratifying.
At the same time, I feel the weight of tremendous responsibility. Popularity does have its price. I am constantly sought out by Epinions members and nonmembers alike concerning items I have reviewed. Many members contact me by e-mail for assistance, which I try to deliver within the extent of my ability and my schedule limitations. I have considered hiding my e-mail address, but to me that would be an abdication of my Advisor role. So I read and rate more reviews than ever, correspond with many members, and provide answers to computer questions and HTML tutorials at my “Computer Teacher” web site. All this while trying to write a review per day and holding down a full time job!
Analyzing Popularity At Epinions
My own newfound status as a “Most Popular Reviewer” gave me pause to think, and I began to wonder: Who are these popular reviewers, and what about them makes them popular? This review is based on my analysis of the publicly available information on the profile pages of the 20 most popular reviewers (according to Epinions’ Community Center posting on October 17 and 18, 2000) at Epinions. By studying these pages and performing some simple descriptive statistical analyses, I was able to generate a composite profile of a popular reviewer and to study some of the variations on that theme.
I visited the Community Center and examined the list of Most Popular Reviewers Overall on October 17, 2000. There were listed 48 reviewers. Of these, I arbitrarily selected the top 20 (in terms of total visits shown on the MPR page). At this point, I should clarify that the information provided on the MPR page is inaccurate. Furthermore, the data shown on the Most Popular Opinions page are inaccurate as well, as was pointed out to me by Computer Advisor MobiProf in an e-mail interchange and subsequently verified by my own visit to the MPO page. Apparently, these pages are not updated on a real-time basis, and once the information is posted, it remains static for some time.
The number of visits shown on the MPR for every reviewer listed there is inflated for some reason. I used the MPR page only to identify the reviewers shown by Epinions as most popular, and extracted the information for member and total visits directly from the member’s public profile page, on the assumption that the profile page data are updated in real time and would be more accurate. I had conducted this analysis on October 17, 2000 only to find out that a server error had inflated the member and total visits on members’ profile pages. I had posted this review, but when I learned that the member and total visits on which the analysis was conducted were inaccurate, I immediately deleted the review--my apologies to those who may have read and rated the earlier version yesterday. My Opinions and Earnings page showed the review as unrated and unvisited when I deleted it, but at least one member (MobiProf) had already seen the posted review. I waited until the inflated totals were corrected by Epinions, and then gathered the current information as of 12:21 P.M. CDT on October 18, 2000. I visited each reviewer’s profile page to document the following information:
• Membership Date
• Number of Reviews Written
• Member Visits
• Total Visits
• Gender of the Reviewer
• Advisor Status
• Number of Members the Reviewer Trusts
• Number of Members who Trust the Reviewer
In some cases, the reviewer’s gender was not discernable from the member name, the photograph, or the profile page. If that was the case, I left that information blank. In most cases, however, either the reviewer displayed his or her gender information, or as in the case of the pulchritudinous forkids, the photograph and the profile left no doubt as to the reviewer’s gender. I indicated Advisor status only if it appeared in the heading of the person’s profile page. I realize that this is not a foolproof method, since some individuals are Advisors in more than one category and since the profile page displays only one (or possibly none) of these. By restricting my information to that available only on the profile page, I may have attributed non-Advisor status incorrectly to one or more Most Popular Reviewers.
A Composite Profile of a Most Popular Reviewer
Using this information, I developed the following composite profile of a Most Popular Reviewer:
An Epinions Most Popular Reviewer is a male Advisor who has been a member of Epinions for 11.1 months. He has written 332 reviews, for an average of almost one review per day. He trusts 138 other Epinions members, and in turn is trusted by 740 others. His reviews have been read by members 29,555 times and have been read a total of 73,778 times.
As in the Presidential polls, the most interesting information is not in the overall results, but in what the pollsters call the “inside numbers.” These break down the responses to examine differences and trends. Accordingly, I examined the ranges in each of the data categories.
The Inside Numbers
• Length of Membership--This ranges from a low of five months to a high of 15 months.
• Reviews Written--These range from a low of 48 to a high of 707 for Movie Advisor Brian Koller. The rate of writing new reviews ranges from a low of 5.8 per month to an astounding 55 per month for the prolific Personal Finance Advisor Profilewriter!
• Demographics--Four women and 14 men, with two unknown--though I suspect both of these are males based on member names.
• Advisor Status--The profile pages indicate that 15 of the top 20 reviewers are Advisors, and as I indicate above, some of the remaining five may be as well.
• Member Visits--Range from a low of 2,246 to a high of 57,758. Interestingly, the reviewer with 2,246 member visits had a total of 72,267 visits for a preponderance of nonmember visits.
• Total Visits--These range from 55,142 to an astonishing 144,203 for Brian Koller. In general, those reviewers who have large numbers of outside visits get those views by writing about topics of public interest such as computers, automobiles, or movies, and use an external partner site to promote nonmember visits.
• Number of Members the Reviewer Trusts--This ranges from a low of 3 other members to a high of 664.
• Number of Members who Trust the Reviewer--This ranges from a low of 68 other members to the amazing 2,334 for ptiemann. No wonder Peter gets 374 HR’s in ten minutes when he writes a new review.
What Do the Numbers Mean?
My interpretation of the numbers is the following. Popular reviewers write well and write frequently, on average one review per day--which has always been my personal goal. Popular reviewers have cultivated a following either in the Epinions community, outside Epinions, or both. This following is indicated by the number of people who trust the reviewer and by the number of member and non-member visits.
Popular reviewers trust wisely. When I exclude the one reviewer who trusts only three others, the ratio of the number of people the most popular reviewers are trusted by to the number of people they trust is about 10 to 1 on average.
Advisor status obviously assists in the popularity of most of these reviewers, but there are notable exceptions. Some very popular reviewers have become popular without the benefit of Advisor status. Advisors or not, popular viewers are persistent and consistent. People who stop writing reviews and stop participating in Epinions altogether will find their popularity waning. Only those who are in Epinions for the long haul can hope someday to make it to the top of their categories or ultimately onto the Most Popular Reviewer page.
Epinions popularity is ultimately based on both the quantity and the quality of the contributions the reviewer makes to the community and its visitors. It is a product both of the person’s individual efforts and his or her willingness to assist others. In the final analysis, one cannot choose to be popular. This is a position into which others place him or her. It was once said that if you ever see a turtle on a fence post, you know it had to have some help getting there. That is the way I feel about my own position at Epinions.
What do you think?
 |
|
|
|
|
|
|
Epinions.com ID:
lap0530
|
 - Top 1000 |
|
Location: Anderson, SC
Reviews written: 122
Trusted by: 332 members
About Me:
Psychology and management professor and business consultant
|
|
|
© 1999-2009 Shopping.com, Inc.