Comments on: Online Video Ranking: MSN/Live Surge, Competitors Sink

By: Frank Sinton

Frank Sinton — Sun, 09 Dec 2007 23:44:54 +0000

Alex-
Thanks for the great summary. I hope Mefeedia makes it on your list in 2008. :)

I have historically found 3rd party analytics understates sites not in the “top 10″. We have been growing very well, yet the numbers don’t always reflect our “source” stats. That said, it is a very useful tool to measure yourself vs. the competition.

Looking forward to more of these reports!

-Frank
CEO, Mefeedia.com

By: Alex Patriquin

Alex Patriquin — Thu, 29 Nov 2007 20:17:13 +0000

Davis,

For now, Compete focuses on the US to ensure the highest standards of data quality. Compete invested significant resources to build the largest, most reliable panel of US internet users in the industry and continues to hold data integrity and transparency as a first priority.

Accordingly any traffic stats are US-based, including YouTube and MySpace. As to your skepticism at YouTube’s size, I suggest comparing the site to Google or Yahoo, which each have >120M UVs. Looking offline, the household comparison is a bit misleading since Compete measures traffic at the individual user level. It might make more sense for you to compare YouTube UVs to the US population (~300M).

Since Compete is a consumer-focused panel that measures traffic at the individual user level, we have minimal work/home overlap.

Hope this helps,

Alex

By: Davis Freeberg

Davis Freeberg — Thu, 29 Nov 2007 18:25:37 +0000

Thanks for helping me to better understand these numbers - That does make more sense. DivX is much more popular in Europe and Asia then it is in the US, so that may help to explain it. Is there anyway for Compete to start tracking international traffic instead of just the US? Does this also mean that the YouTube, MySpace and other numbers don’t reflect the traffic from outside the US? 50 million uniques in the US alone would mean that about 50% of all US households went to YouTube last month. I know that the site is popular, but this seems a little high to me. I’d also be interested in knowing if Compete has any data on how much overlap there might be with some of the uniques. If I visit a site at home and at work, that obviously would count as two uniques. It’s probably hard to get at this data, but I’d be interested in knowing if you have a sense of what percentage of uniques may be counted more than once. This would obviously change from site to site, but it’d be interesting to see a ballpark figure for the overlap of work/home visits on the net in general.

By: Alex Patriquin

Alex Patriquin — Thu, 29 Nov 2007 16:10:30 +0000

Davis,

DivX shows up just shy of the Top 10 in this month’s ranking. Compete.com’s Site Analytics covers the entire domain, so the subdomain “stage6.divx.com” is included in the UV count for divx.com that you mentioned.

There are several possible explanations for the discrepancy.

1) Cookie deletion - If DivX is citing a 1st party web analytics product (like Omniture, Google Analytics, ect.), those products measure visitors by counting cookies. Since visitors often delete their cookies multiple times in a month, they could be overcounting by a significant margin.

Compete does not count cookies. For more on our methodology, see http://www.compete.com/help#snp1.

2) Compete traffic statistics are US-based. This could be a factor if DivX has a large international audience.

3) I would strongly caution against using Alexa traffic data, but don’t take my word for it. There is plenty of commentary on Alexa these days. I recommend reading up and deciding for yourself.

It is important to note that you should consider any web analytics in context. Generally, it’s not very productive to compare 1st party analytics (Omniture, Google Analytics, etc.) with 3rd party (Compete) because of the cookie deletion issue and because they are not true substitutes.

Where 1st party analytics go deep on a site’s internal traffic, 3rd party analytics compare traffic across different sites. As a result, these two types of services have drastically different measurement methodologies and often produce widely discrepant results.

There may be some other possible reasons for the DivX discrepancy that I’m not aware of. Usually Compete consults with our corporate clients to discover and control for these discrepancies.

Hope this helps,

Alex

By: Davis Freeberg

Davis Freeberg — Wed, 28 Nov 2007 16:10:27 +0000

What I don’t understand about your results is why DivX’s Stage6 video site isn’t showing up on this list. According to Compete, they get less then 1 million visitors per month, but according to DivX, they are averaging over 10 million uniques. I thought that the discrepancy might be because Stage6 is a sub domain of DivX.com, but even when I check the stats for DivX.com, it shows less then a million uniques. Can you help me to better understand why Stage6 doesn’t show up in Compete’s filters and some of the reasons why the traffic difference would be so great. I could understand if it was off by a couple of million visitors, but on Alexa, they Stage6 as a top 500 site.