Analysis: What are the Web's Top Sources of Referral Traffic?
If there's one thing we know about Web authors it's that they are constantly seeking new sources of traffic for their content. It doesn't matter if you're a blogger, a marketing manager or a small business owner, there is simply no reason to invest time with content creation and Web design if no one is coming to read it. For this reason, it's important to figure out where to actually invest time for the greatest ROI.
As a blogger I routinely asked myself the same question, until I finally realized that I (as the CEO of Woopra, the Web analytics company) had access to all the data I needed to make an absolute determination about which areas deserve the most attention.
Guest author John Pozadzides is the CEO of the Web analytics company Woopra, organizer of the open-source blogger and developer conference OpenCa.mp, and the man behind OneMansBlog.com. Follow him on Twitter @johnpoz.
This meant looking beyond my own site's statistics and finding out what was broadly occurring across all domains. Only through a comparative analysis can we determine where our strengths and weaknesses lie as compared to the average.
The Methodology
Woopra monitors well over 100,000 websites, however, all of our client's data is kept separate for security and privacy reasons. This meant the first step was to set up a new specialized server that would poll all other servers in our network in order to aggregate the data.
After building this new platform, the Woopra team provided me with the raw information that I used to create the charts below. This report is based on hundreds of millions of data points collected during the month of June 2010.
Virtually every category of Website is represented in the dataset - education, news, government, SMB, Fortune 500, blogs, adult - and while it clearly doesn't reflect all of the sites on the Internet, my guess is that it is a statistically significant and representative sampling.
Top Referrers of Traffic
For the purposes of this analysis, referrers have been segmented into different categories in order to more easily compare traffic. For example, there is no use in comparing Google to Flickr since they are not categorically related. If we're making a determination which search engine to focus on, Flickr would not be in the mix - and if we're looking for a photo hosting site, Google would not be in the mix. The four main referral categories that drive virtually all traffic are: Search Engines, Media, Social Bookmarks and Social Networks.
Social Network Referrers
Perhaps the one result that was the most surprising is the Social Networking category. Although Twitter seem to be the talk of the town, Facebook is the 900 pound gorilla when it comes to actually driving website traffic, sending nearly 7 in 10 visitors from the Social Network category. LinkedIn comes in a distant yet still respectable third place.
There is one giant unknown in this area, however. Many Twitter users access the service through applications instead of via Twitter's website. These applications do not report http-referrer data to Web servers, which makes it impossible to tell where the clicks originate. The same can be said about traffic driven by mobile apps (including Facebook).
All of the other players in the Social Networking category cumulatively amount to less than 3% of inbound website traffic.
Social Bookmark Referrers
Social bookmarking sites are also extremely sought after sources of traffic, with the highest profile member of the group being Digg. Interestingly, this is another case where a smaller source seems to receive a disproportionate amount of attention, since StumbleUpon drives nearly double the traffic to websites.
Another surprising standout in the Social Bookmark category is YCombinator's Hacker News site, which drives 12% of the traffic in this category. Reddit and Del.icio.us drive 5% and 2%, respectively. Interestingly, SlashDot (the grandaddy of them all) drives close to 0% of traffic to the sites Woopra measures - although webmasters still call the rush of traffic that can take down a website a "Slashdotting".
Search Referrers
In the Search Engine space, perhaps the only surprise is the absolute dominance of Google when it comes to actually driving traffic to websites. For example, Experian's Hitwise published a press release setting Google's share of the search market at 72% in May 2010. However, Woopra is seeing 92% of search visitors originating from Google.
This begs the question: Where does the discrepancy lie?
- Is the Hitwise data set more representative of the whole?
- Are there more searches taking place on other search engines that don't result in click-throughs to websites?
- StatCounter's Global Stats seem to closely mirror Woopra's data.
Regardless of the reason, what we do know is that focusing attention on optimizing for Google search is absolutely the right thing to do. (Here's the official Google SEO starter guide.)
Media Referrers
The media referrer category is reserved for sites that focus on various forms of multimedia, including images, video and audio. These sites are often overlooked as a source of traffic because they are destination sites; however, media sites can also drive large volumes of traffic.
As an example, world-famous HDR photographer Trey Ratcliff posts travel photos on Flickr, and includes a simple link back to his site in the description of each image he posts.
This technique generated nearly 13,000 pageviews on StuckInCustoms.com in April alone.
What's more notable is that Flickr is not the dominant referrer in the Media category. YouTube drives 900% more traffic to websites than Flickr. This is accomplished the same way, via links in the description of the video.
Amazingly, 99% of referrer traffic in the Media category comes from only four providers: YouTube, Flickr, Last.FM and Vimeo.
Lessons Learned
The only question we are trying to answer with this analysis is where it makes sense to focus resources and attention in order to drive traffic. So here's what we know:
- Search engine optimization specifically targeting Google's index is key. Working hard to satisfy other search providers probably won't pay off as much as spending the time elsewhere - like Facebook and Twitter promotion.
- If multimedia isn't part of a site's strategy to drive traffic, it should be. Adding photos to Flickr and videos to YouTube will not only pay off in terms of adding content for your brand, but will also drive eyeballs to your site.
- Sites that are not seeing traffic from Facebook, Twitter, or both, are missing out on a major traffic-generation source. Additionally, if the balance is not in Facebook's favor, you are likely under-utilizing this channel.
- Don't waste time promoting your site on any Social Media network other than Facebook, Twitter or LinkedIn. Your time is better spent elsewhere (like YouTube or Flickr for example).
- By all means, submit articles to Digg in the hopes of making it to the homepage, but more importantly make sure articles work their way into StumbleUpon, which can bring a larger, more sustained and less server-crash-creating stream of traffic.
- For those looking to more accurately track Twitter click-throughs, some service providers such as Bit.ly enable tracking of short links. The bad news is that the tracking is not aggregated into existing Web analytics provider data.
- This analysis doesn't take into consideration back-links (which Web authors should pursue whenever possible) because on the aggregate they do not drive global traffic. However, links from related or high-traffic sites can not only dramatically increase site traffic, they can positively impact search result listings.
Questions, comments or alternative interpretations? Drop a comment below and let us know!
Top photo by Lars Sundstrom
Share
Facebook
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Love it--I'm sure a lot of people are about to adjust their marketing plans based on this.
Twitter and Facebook clients should report themselves as such. It would be interesting to see direct mobile browser traffic, I bet a lot of it actually comes from Twitter clients.
Can you cite your sources on those percentages? I'm curious to know if this was based on data for a single site, or if this was many different sites. Also, were the visitors to the sites that were tracked normally more socially active on websites.
Thanks
Sean,
As I stated in the article:
I believe that our sample size was large enough to be statistically significant, so the visitors were probably representative of the general whole of the Internet.
I believe that another stats company would have to publish a similar report to determine if they are seeing things that are different than we do, but my informal research indicates that its pretty average.
Cheers,
John P.
I too believe that, based on your methodology, this is a statistically significant, representative sample. Another quality post from "the man" John P.
I just signed up for Stumbleupon and see why it's such a traffic generator. I'm going to have to figure out how to get it working for me.
Can you provide any insight into the overall percentage of web traffic generated by each of the 4 categories? Search Engines, Media, Social Bookmarks and Social Networks. I am curious to know whether Search Engines dominate overall, or whether Social Networks may approach a similar volume of referrals?
Thanks for the insightful report!
Using last years numbers, there were around 234 million websites on the web in Dec-2009. Guess we could easily take it to 250mi as of today, right?
So, if we account that your 100k+ records are a true "sample" from the web, in a perfect world that would mean that your research has an error margin of ~0,31% wich i assume is VERY good! =)
On a sidenote, it would be VERY interesting if you could break down the demographics and base-content of those sites (I could help you if you don't have time.. lol)
I am very interested on e.g.: "what are the top referrers for websites of charity/non-profitable-organizations?" (I am a director of one).
Still, many thanks for the report! =)
Super-useful article - thanks for pouring over the data and bringing us these insights!
Would like to add one caveat to all this. This report deals only with quantity of traffic - not quality. To take this one step further, we'd need to know what type of traffic creates the conversions that are significant for each site. But overall - awesome information!
PS: I took my numbers from Pingdom's research "Internet 2009 in numbers" and the stats were calculated using a statistical tecnical book as reference.
The question is: Do these sites promote their pages equally on these services?
Guess that a lot of sites only or mainly use Facebook to spread their sites and news. So of course most links are coming from there.
I spoke to several smaller web site owners. The ones who spread on Facebook and Twitter all stated that Twitter is driving a lot more traffic than Facebook.
And just look at Mashable and TechCrunch and the numbers how often they are liked or retweeted. OK, tech people maybe are more using Twitter than normal people. But I remember a study that Twitter users are more likely to click links than users of other social sites.
Think this statistics is not very useful as long as it is not clear how and where these sites link and spread their links!
I believe this is really useful information.
I agree with most of the above information & stats and would like to share my personal experience that no doubt, google is dominating in search engine race. The recent stats from one of my Blogs showed google=67%, yahoo=3.3% and Bing=2.5%.
Apart from search engines other referral sources also depends upon the nature of your site/blog - Educational.Photography, informational, article based and so on.
You cannot get the same results for different type of sites/blogs. You also need to carefully evaluate the type/category of the blog/site and focus on related media/source to generate traffic.
But mostly other good sources are Facebook, twitter, youtube and social bookmarking sites.
There's a huge elephant in the room here: Blogs, news websites, email, message boards? You can't tell me that these all don't send a substantial amount of traffic.
Very useful article. May I know the sources of data used for the analysis?
Didn't know that one can put links to flickr photostream page. Never heard of Ycombinator before this. StumbleUpon's domination is impressive. Have to learn how to use it properly.
Great article, but is referral traffic even the right metric to worry about?
Optimizing the wrong metric can be the root of all evil for marketing:
http://smoothspan.wordpress.com/2010/07/29/optimizing-the-wrong-metric-may-be-the-root-of-all-evil-in-marketing/
Cheers,
BW
Where is the pie chart showing the absolute size of each segment of referrers?
very elaborate and nicely put. I guess the primary source for man blogs would still be bookmarking sites like Stumbleupon,Digg etc.
I need to get my act together on Facebook ASAP. Great stats, thanks.
This is freaking educative,i cant but tell you that i have learned alot..wow....Thanks
@ John
Great article!
I'm an early fan and -user of Woopra. Since I was granted access as a Beta tester I have had a second Woopra dedicated screen next to my main screen to monitor my 3 niche blogs real time.
The findings are in line with my own, although the Youtube media referral part is an eye opener or maybe a bit overstated. I wonder why you did not consider to take Google image (and maybe Bing image referrals as well) referrals into that equation as media referrals. It is my impression that Google image searches do constitute between 40 to 60 % of Google's total referrals. It might me that the Youtube referral part is overstated...on the other hand I have heard a Google guy stating that youtube is the number 2 search engine..
More observations:
Facebook's recent overhaul and ease of use contributes more and more to referrals recently.
The recent new "Bingish" way of presenting image search results by Google may influence their referrals negatively as the algoritme is more "caffeinish" i.e. more adhoc. I've a feeling they are letting the ball drop in the same way as Technorati dropped the ball, but I may be wrong.
Flickr is notoriously behind in ease of use when compared with Facebook now and that may explain their relatively low part in referrals. On the other hand Google image search results do include Flickr photos, so it is not easy to establish Flickr's part.
My .. cents
Great Article, Now I'm going to have to change my entire marketing plan! Thanks a lot! Great article and genuinly useful info. I was just amazed by the stranglehold google has on the entire internet!
This is very useful data.
One question: Where are most of Woopra's customers located and where are the customers they target?
What pushes me to ask this is your chart showing search referrers. It doesn't include major Asian search engines Baidu and Naver. If most of your clients are in the U.S. and Europe, there may be some source bias.
I commented already yesterday, but got an idea for explanation why StumbleUpon is so huge in its category. Maybe it is because of their PPC advertising program?
http://www.stumbleupon.com/ads/
Love the article- very powerful and it will drive many marketing plans including my own- very nice
"Experian's Hitwise published a press release setting Google's share of the search market at 72% in May 2010. However, Woopra is seeing 92% of search visitors originating from Google.
...
This begs the question: Where does the discrepancy lie?
Is the Hitwise data set more representative of the whole?
"
From the looks of the link you posted, the 72% is for Google.com US only, and Hitwise also only tracks US traffic. Whereas it seems you are including all international google properties and presumably Woopra's data is not restricted to US traffic only?
Excellent article. I have taken away some gold nuggets from reading this. Really helps with all the graphs, and then thoughtfully summed up in the conclusion. I'm putting this in my resource list to refer to at a later date.
These are great tools, i personally use clicktale which has the best of the bunch in my opinion. Its heatmaps are second to none.
I really wonder to see facebook ahead of twitter as, twitter seems to be much flexible to share tweets and following and unfollowing procedures are very less time consuming. After reading your post I would like to try some hand on facebook now.
It will be interesting to re-look at these stats if Facebook do manage to get their own search sorted. It could potentially level out the playing field a bit more with Google if they do manage to get their own search algo running well!
@Bob you're the only one that gets it around here.
It's not about number of referrals. Please. I could care less if 10,000 people from Reddit visit my site if it results in ZERO profit. It looks nice in my graphs and stats, but it doesn't gimme money.
For most ppl, marketing in Facebook and Twitter is a waste of time. It's all about their attitude. They're not in a buying mentality. But someone who is searching for "buy tax software" or "best online backup" is.
Where is the pie chart showing the absolute size of each segment of referrers?
http://mobile8cz.blogspot.com
Providing a community to share files, photos, videos, music, friends, games, online chat social network and more at http://enjoy2web.com.
This is the kind of solid, research based data we need to justify our decisions to the ultimate decision makers. Thanks, John P.
Brad