ProBlogger - Latest Posts

Search, Social and Direct Traffic - [TRAFFIC ANALYSIS]

Posted: 17 Nov 2008 06:03 AM PST

This morning I spent a little time doing some analysis (using Google Analytics) of the traffic coming into my main blog - Digital Photography School.

My analysis was stimulated by a question from a reader who in response to last week’s two posts examining the place of Digg and Social Bookmarkingin a bloggers priorities asked me:

What role does Social Bookmarking traffic play in your blog?

I decided to dig into the metrics on DPS and find out the answer… or at least that is what I started out doing…..

As I began to analyze the stats I realized that DPS has four main referrers of traffic - each are quite different from the others and yet each are very important. What follows in this post is me thinking out loud on each source of traffic and what it means to my blog.

Looking at the big picture

Lets start by looking at the big picture of the traffic coming into DPS. For the purpose of this post I’ll go back to the start of 2007 with my analysis (the time I started using Google Analytics) and I will only be looking at traffic coming into the DPS blog (ie this doesn’t include data on the forums).

Here’s a snapshot of all traffic coming into the DPS blog since 1 January 2007 (click to enlarge all images in this post).

You can see over the last 22 months that the DPS blog has had steady growth. There have been 11.5 million visitors, around 25 million page views and they stay on the site around two and a half minutes per visit.

At 1 January the average daily visitor numbers were around 4,000-5,000 visitors. At present they average around 23,000-25,000.

Looking specifically at the main sources of traffic to the blog - there are four that are responsible for a little under 70% of all of the above traffic:

Google (26%)
Direct Traffic (RSS, Newsletters, Browser Bookmarks etc) (21%)
StumbleUpon (11%)
Digg (9%)

The next highest referrers are significantly lower in how much traffic they bring in and include Yahoo, many other blogs (big and small) and Delicious.

As you can see - Google is a fairly important factor in my blog. Add other search traffic from Yahoo, MSN, AOL and search traffic is responsible for around 30% of the overall traffic.

If I was to categorize all of the social bookmarking traffic (Digg, StumbleUpon, Delicious, Reddit, Popurls etc it accounts for around 24% of overall traffic (a little higher than ‘direct’).

OK - so this information is mildly interesting (to me at least) but when I dig down a little further and do some analysis of each type of traffic I find it more illuminating.

Digg Traffic

Since last week we were talking about Digg, lets start with that.

Here’s how Digg traffic to the DPS blog has looked over the last 22 months.

Straight away we can see the nature of Digg traffic. It is either there or it isn’t. The spikes can be fairly significant (in most cases they range from 10,000 to 30,000 visitors) but between them the traffic from Digg rarely gets over 100 visitors a day.

Lets look at a few other stats on Digg visitors over this period:

They viewed 1.39 pages per visit (site average was 2.17)
They spent an average of 54 seconds on the site (site average was 2 minutes and 35 seconds)

So in comparison to overall averages Digg users are fairly fleeting (although note as fleeting as I hear some people saying).

One other thing worth saying about Digg visitors. I’ve heard a lot of people say that they don’t ‘convert’ to regular readers. So lets have a look at my newsletter signups for the latest ‘Digg Event’ on DPS (that last spike on the chart).

As you’ll see there was a definite increase in subscriber numbers on the day of my last Digg event (Nov 13th). Of course that day had 14,000 visitors from Digg to the site and subscriber numbers were only up around 150 subscribers - so Digg users don’t become loyal readers in huge numbers - but some of them do convert. I’d suspect that RSS subscribers would increase by a similar sort of rate after a Digg event.

I’ve noticed similar sorts of increases in subscriber numbers on other ‘Digg events’. They don’t convert massively but I always do pick up extra readers each time - the stats on the site tend to look like this chart taken from my post - How to Build a ‘Digg Culture’ on your Blog:

200612091300

This is actually one of the biggest benefits of social bookmarking traffic for me. While the actual spike in traffic is nice - the real benefit comes from those readers you’re able to convert to regular readers. 100 extra readers adds up to thousands of page views over a year.

One more stat on ‘conversion to loyalty’:

Over the last few months I’ve had a test running on Google Analtyics that analyzes how many visitors ‘convert’ to subscribers. I’ve set up a ‘Goal’ on Google Analytics that is triggered as achieved when people reach the thank you page for my newsletter subscription (meaning when they convert to verified subscribers).

Digg Users get to this page 0.48% of the time. This is in comparison to an average of 2.24% for the overall site.

Do Digg Users Click Ads?

One of the great things about Google Analytics now is that you can track AdSense earnings if you link your AdSense and Analytics accounts (they’re still rolling this feature for some).

While AdSense TOS prohibits sharing of too much information on earnings I’ll share some vague stats with you on how different readers ‘convert’ with ads.

The CPM (earnings per 1000 page views) has converted with Digg readers at about half the site average.
The CTR (click through rate) of Digg users is about a third of the site average.

So the common perception that Digg users don’t click ads is backed up - to a point. Some of them do click and when you consider that you can get 30,000 of them visiting your site in a day this can add up.

Keep in mind that Digg traffic can be useful for monetizing a site in other ways - particularly when you’re making money on a CPM basis where you’re paid per page view.

StumbleUpon Traffic

StumbleUpon actually sends me more traffic than Digg does over time. Here’s how the traffic from SU looks over the last 22 months.

Here we see that the nature of Stumble Upon traffic is actually quite different from Digg. While both are ‘bookmarking’ sites they are really quite different. When a post gets popular on StumbleUpon the traffic it generates is spread out over days (and even weeks and months). There’s often no single day when you get masses of traffic but rather it’s more of a slow burner (I’ve written more about this in a post titled Why StumbleUpon Sends More Traffic than Digg).

You’ll see that StumbleUpon traffic has actually grown significantly over time. What I put this down to is that as I’ve written more and more posts on my blog there have been more entry points for SU traffic. While traffic grows and then falls off to particular posts on SU if you have multiple posts generating traffic you can actually see it build to significant numbers (like they were in the period of June/July this year where I had about 6-7 posts doing very well in SU simultaneously).

Lets look at a couple of other metrics on the SU traffic:

They viewed 1.62 pages per visit (site average was 2.17)
They spent an average of 1 minute and 7 seconds on the site (site average was 2 minutes and 35 seconds)

So StumbleUpon traffic is a little more sticky than Digg traffic. They view more pages and stick around longer.

Do StumbleUpon users signup for the newsletter and become loyal? My stats show that 0.51% of them have reached the thank you page on my newsletter subscription process. Slightly higher than Digg users but a lot lower than overall site averages.

Do StumbleUpon users click ads?

Interestingly StumbleUpon users seem to click on ads less than Digg users with the limited amount of stats that I have on this. The CPM that I’m seeing with SU users is very similar to that for Digg users but the CTR was about a third of Digg users (and about a tenth of overall site averages).

Search Engine Traffic

My number one traffic source on DPS is that from search engines. Google takes the lions share of this but I’ve added in the others into this analysis (interestingly Yahoo has been on the increase of late). Here’s how the search engine traffic has grown over the last 22 months.

Again - a very different shaped chart to the others. The two spikes in traffic are both to do with search traffic increasing for terms around ‘fireworks photography’ at around 4th July - but other than that it’s very steady growth with little weekly spikes and troughs in traffic but not much else to note.

This traffic has gone up over time for a couple of main reasons:

1. I’ve been adding content - the more pages you have the more entry points that search engines can send people to

2. The sites authority has grown over time - the longer you’re around the more links you have pointing at your blog and the more authoritative search engines begin to give you.

Lets look at a couple of other stats from Search Engine Traffic:

They viewed 2.55 pages per visit (site average was 2.17)
They spent an average of 3 minutes and 20 seconds on the site (site average was 2 minutes and 35 seconds)

Interestingly Google readers view 2.51 pages and spend 3 minutes and 16 seconds while Yahoo readers view over 3 pages and spend over 4 minutes on the site.

In terms of ‘conversion’ via the newsletter - 2.72% of search engine visitors have made it to the thank you page (again it’s better for Yahoo than Google). This is better than the site average making search traffic more sticky than social media traffic.

Do Search Engine Readers Click Ads?

The common perception is that search engine referrals are more profitable when it comes to CPC advertising programs like AdSense. My stats back this up.

I’m seeing the CPM of my search traffic as about 10% higher than the site average and CTR up by about 10% also. Interestingly I’m seeing Yahoo traffic as about 30% higher than Google.

Direct Traffic

The last category of traffic that I want to analyze is what Google Analytics classifies as ‘direct’ traffic. This traffic includes those coming in from desktop RSS subscribers, newsletters, browser bookmarks, type in traffic etc. Here’s how this traffic has looked over the last 22 months.

Again we see a fairly steady growth in this area. The weekly spikes coincide with when I’ve sent out newsletters. The bigger spikes mainly coincide with when we’ve run competitions in our newsletters.

The reason for the growth in this traffic is largely that I’ve worked very hard on building a newsletter list for this blog (particularly over the last year).

Lets look at some more stats on this direct traffic:

They viewed 2.28 pages per visit (site average was 2.17)
They spent an average of 2 minutes and 55 seconds on the site (site average was 2 minutes and 35 seconds)

Both of these stats are higher than the site average but lower than search engine traffic. However considering that many of these visitors come to the site on a weekly basis and view hundreds of pages a year these averages are pretty good.

In terms of ‘goal conversion’ (or getting these people to my thank you page of the newsletter signup - they convert at 2.08%. This is slightly under the site average but considering many of them have already signed up - it’s pretty good.

Do Direct Referrals Click Ads?

This one interested me because I suspected that these highly loyal readers would become pretty blind to AdSense ads over time. However they are bang on average for the site with both CTR and CPM performance almost exactly on the site average.

Concluding Thoughts

I know this post has been rather long and so I will keep my concluding thoughts brief (I considered posting this as a series of posts but hope it’s more helpful seeing everything side by side).

All traffic has its place and serves different purposes.

One of the main things that strikes me about this exercise is that while some people write off different types of traffic - that together they come together in fairly significant ways.

For example - Digg traffic may not be that sticky or profitable - however as I think back to the early days of DPS it was the early series of Digg spikes that helped to get the blog going.

Even going back before January 2007 (before the charts above) DPS was on the front page of Digg quite a few times. Each time this happened the site step ups in loyal readers to the blog. This helped it grow even though at the time the site wasn’t generating much search traffic.

Overtime search has been increasingly important to the site in finding new visitors. The Digg spikes are handy and still draw people in that have not seen us before but in many ways they’ve served their purpose for the site and now our Google and Yahoo authority has kicked in we’re starting to see more benefits from there.

As I look forward I see both ’search’ and ‘direct’ traffic as taking over even more from social bookmarking traffic. If things continue to grow as they are search and direct traffic will out number even the biggest spikes that the site might get from Digg.

This doesn’t mean I’ll not value the bookmarking traffic - but it’ll play less of a roll.

Social Bookmarking as an SEO tool

One last unproven idea that has been lingering in my mind lately is the importance of social bookmarking as an SEO strategy. I’m not sure how much of an impact it has had on the growth of search traffic on DPS but surely all of the links to DPS from Digg, StumbleUpon, Delicious, Reddit and other social bookmarking sites have had an impact upon the site’s search authority.

Even posts that don’t get to the front page of Digg that are bookmarked there must at least be getting some search engine juice from the bookmark.

More than that - getting on the front page of Digg or going popular on Delicious often has the flow on effect of being linked to by a lot of other blogs and websites that watch these pages. For example my last appearance on the popular page on Delicious stimulated at least 30 or so links from other blogs. Again - each link is adding to the search engine authority of the blog.