Analyzing Trends From Seth Godin's Web 2.0 Index
By Alex Iskold
Marketing guru and blogger Seth Godin is also known for his Web 2.0 Traffic Watch List on Alexaholic. This list tracks the changes in Alexa traffic for about 1000 Web 2.0 companies. For example, MySpace and YouTube are numbers 1 and 2 respectively (although check our previous post, which has YouTube in the top spot). While there has been a lot of skepticism about the precision of Alexa ranking, particularly because the ratings can be bought or gamed, for a large pool of popular web sites it can adequately be used as a relative measure of popularity.
Here is the latest Seth Godin Alexaholic top 10 list:
The list has generated attention in the blogging community. Paul Kedrosky wrote a post recently entitled Hot and Not So Hot Web 2.0 Companies, in which he identified the biggest web 2.0 gainers and losers over the last 6 months. We found this interesting and wondered what else can we deduce from Seth's chart. Instead of individual companies, we wanted to determine what trends are gaining or losing. For example, is online video continuing to rise? Does social bookmarking have a chance to go mainstream? We looked into Seth's data to answer these questions.
How we looked at the data
Seth Godin's list shows only companies, but not trends. As the first step, we labeled each company with a category such as Video or Social Networking. Next, we calculated additional information needed to determine growth. For each company, we calculated the % Change in Rank. This is the number which indicates relative change in the ranking of each company. The reason for using this percentage change instead of the actual change number is to take into consideration that fluctuations at the top of the list are much smaller than fluctuations at the bottom. Here are the first 10 entries from the new list:
Trends for top 100 companies
We then labeled the top 100 companies with the categories and calculated the common values for each category. Here is what came out:
Green = Growth; Red = Decrease
Important Note: lower figures mean more popularity
There is a lot of interesting information. First note that Photo is the most populated category in the top 100, with 10 companies falling under this category. It is closely followed by other usual suspects: Search, Video and Social Networking.
Looking at the Average Rank column gives us the current average standings. Since lower Alexa ranking implies more popularity, Photo and Music hold the lead, followed by Blogging and File Sharing (note that File Sharing includes storage, like Box.net). Among the worst rankings we find Search and Video. Search is probably is not surprising, because Google has a strong grip on this category. The reason that the Video average is not great is because the players in the top 100 are at the bottom, with the exception of course of YouTube and DailyMotion.
Perhaps the most interesting information is the Average % Rank Change, because this information indicates growth. Remember that the lower the figure, the more growth it indicates. We were somewhat surprised to find that the six players in the Social Bookmarking space saw the most substantial growth in the last six months. This indicates that social bookmarking is heading towards the mainstream. It is not surprising to see big growth in video, but interesting to see Meebo contributing strongly alongside Skype to the Communications category.
The data also reveals more bad news for the Search category, as it experienced an overall drop in the last six months of 3% points. Again, fighting Google is no easy thing. Finally, the Blogging category had the biggest drop of 9% points. Perhaps this can be attributed to people blogging on their social networking sites, but this is not obvious - although Xanga, Livejournal and SixApart all decreased in popularity. It could also be that more people are using Wordpress, which is on the rise (but not on the list).
Conclusion
It would be interesting to expand on this research and look at the trends for the entire 1000 companies. It would also be interesting to see what companies are influencing the trends the most. Surely the MySpaces and YouTubes of the world, but what other companies? The information on Seth's list can be analyzed in a number of different ways, to gain insights into what is happening in our fast changing Web 2.0 world.
You are welcome to use the spreadsheet we created to do more analysis. Of course the only catch is that you have to share the results with us :-)
Update: There was a slight error in the original spreadsheet, so we have amended that. Thanks Dermot for pointing it out.
3 TrackBacks
TrackBack URL for this entry: http://www.readwriteweb.com/cgi-bin/mt/mt-tb.cgi/2940
FOLLOW RWW ON TWITTER
RECENT JOBS
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
It's hard to say that the average rank for search-related sites is 2902 because this rank excludes market leaders Yahoo and Google (because Seth's list doesn't include them). Whereas, Social Network leaders MySpace and Facebook are included... kinda odd.
Posted by: Andrew Parker | January 10, 2007 12:44 PM
Andrew,
Yes, this analysis only applies to the data in the list. Seth does not consider Yahoo and Google to be Web 2.0.
Alex
Posted by: Alex Iskold | January 10, 2007 12:46 PM
Fun analysis and thanks for doing it. I wish you had access to better data then alexa, though! Someone should hook you guys up with some comscore or hitwise data!!
Posted by: Ted | January 10, 2007 12:58 PM
Ted,
Great idea! The spreadsheet that we provided is generic so anyone can plug in other data with a bit more work.
Alex
Posted by: Alex Iskold | January 10, 2007 1:00 PM
Alex:
This is a really interesting analysis. However, I think it's important to point out some of the limitations of the underlying Alexa data, that qualify the results you've arrived at (which is not your fault!):
1. Only specific companies are included in the list, as Andrew has noted above - e.g. Google and Yahoo are excluded.
2. Alexa's data is suspect, to put it mildly. I understand the "well, it's inaccurate, but we're looking at it in a relative way" argument, but I'm skeptical. Getting good relative results from inaccurate data is problematic at best; for example, there could easily be a correlation between Alexa toolbar users and specific sites (based on, say, a common variable like geography), which would heavily skew the results.
3. Alexa says that its ranking results are based on a "mixture of page views and user reach". Page views is a metric that is increasingly being questioned, in the Web 2.0/Ajax/componentized world.
[An interesting twist for Search engines: a better search engine would find results faster (i.e. with fewer user requests) for a given search, which would result in a smaller number of page views for the same number of users, so that engine's numbers would actually get penalized for getting more relevant results! Do you agree?]
4. As you mentioned, Alexa results are being targeted for gaming or manipulation, which also subverts the "relative rankings" argument.
5. Alexa does not consider "https" results at all. This may have an unequal impact if some sites have more https pages than others (e.g. if they have slightly different offerings).
Having said all that, I think this may be as good a way to analyze relative rankings as any. Compete and Quantcast probably suffer from similar issues (Compete has a comparison with Alexa on their site). I have been thinking about ways to evaluate the buzz-worthiness of a site; some other options are:
a. Number of del.icio.us or other bookmarks [suffers from adoption skew of delicious?]
b. Number of blog posts about the site in technorati or some other blog search engine [ditto adoption by bloggers?]
c. Advertising prices [not public information, but in the long run, the best metric? :-) Trust the free market!]
Posted by: NitinK | January 10, 2007 1:18 PM
Nitin,
Thanks for this detailed comment. I agree that Alexa has many flaws. The reason for my post is to have a different look at Seth's data, thats all. You can apply this analysis to any data, the real focus is on delta and trends.
Alex
Posted by: Alex Iskold | January 10, 2007 1:33 PM
Great reading, thanks. I like it, a differences look at the data.
Posted by: Sendar | January 10, 2007 2:33 PM
Hi, I enjoyed this post, but you should not use red and green to color-code graphs: 25% of males are red/green duotone color-blind!
Posted by: Nick Miller | January 11, 2007 9:26 PM
Good point Nick - thanks!
What do you recommend in this case? Black / White?
Alex
Posted by: Alex Iskold | January 11, 2007 9:35 PM
Very interesting stuff. I've been doing a comparison of Alexa traffic rankings for the leading on demand CRM vendors over the past six months which a few people have found of interest. More so from a relative rank standpoint than believing the rank is totally accurate.
Posted by: Brent Leary | January 12, 2007 8:16 AM
Check out the new embedded version of Kool IM.
www.koolim.com
Posted by: amanda | January 23, 2007 8:29 AM