Google Analytics: referring sites
Google Analytics (the free website statistics package) can tell you a lot about your website, and one of those things is where your visitors are arriving from. This is where knowledge of your ‘referring sites’ proves useful.
Here’s a snapshot of my web stats, from 23rd July – 22nd August 2007.
Interpreting your web stats
It might not look like much, but I can learn some useful things from this info. First things first, a big thanks to Vitaly and Sven at Smashing Magazine, Darren at ProBlogger and Tara at Graphic Design Blog. I’ve received some great traffic from your sites this past month, and it’s very much appreciated.
Okay, on to the stats. Top of the list, for traffic referrers, is StumbleUpon. Who here isn’t familiar with StumbleUpon? It has over 3,000,000 members, it’s free, and as you can see from my stats, is incredibly powerful at attracting traffic to your website.
I think a lot of people under-value the importance of StumbleUpon, and Muhammad at Pronet Advertising recently wrote about the unappreciated StumbleUpon effect:
According to eBizMBA, StumbleUpon ranked at number 6 for the 30 most popular social bookmarking sites for July 2007. The site was ranked above both Slashdot and Reddit but still is one of the least appreciated sources of traffic.
The downside of StumbleUpon traffic is that the average visitor doesn’t spend very long browsing your pages. Sometimes they’re gone before you can say, “Wait, don’t lea…”.
Google Analytics shows how long, on average, visitors spend browsing your site, and the table above clearly shows that StumbleUpon traffic spends the least amount of time on my site out of the top 10 referrers. This is because ‘Stumblers’ in general, prefer mainly image based, web pages - things to look at rather than read, and you can view a photo a lot faster than you can read an article.
Maki at Dosh Dosh has written a very comprehensive guide to StumbleUpon, and how you can leverage it to attract masses of traffic. Certainly worth a read, and he says a lot more about it than I can.
Don’t underestimate the value of your website’s images. Number 3 on my list of web traffic referrers is Google’s Image Search. People are visitng my website having found it when searching Google for specific images. When inserting an image into a blog post, be sure to label it with an ALT tag. I’ve added alt=”Google Analytics David Airey” inside the image code brackets for the table above. ALT tags help search engines read what your images are about. Search engines don’t see the picture, only the code.
Patrick at The Lonely Marketer recently wrote about image marketing, and gave these tips:
- Alternative (alt) text is what shows when the image does not show. This alt text is also crawled by the search engines. Make sure popular keywords are included here.
- Relevant keywords should be in text around the image.
- Anchor text (hyperlinked text) leading to the page and image should include relevant text about the page and image. Instead of “Pictures Here”, try something like, “See photos of SES San Jose and the Google Dance”.
- Photos should be in .jpg or .gif, but .jpg is the best.
- Utilize caption or label text in the immediate area around an image.
- Make sure file names make sense to the audience. Rather than “img002.jpg”, try “patrickschaber-sessanjose.jpg”.
- Use dashes and not underscores in filename.
Another of my top 10 referrers that deserves a mention, is Wikipedia, storming in at number 4. Since Wikipedia linked to my logo designer article, back in June, I’ve been consistently receiving high quality traffic day in, day out. Why do I call it ‘high quality traffic’? If you look again at the table, you’ll notice that visitors from Wikipedia browse, on average, 3.61 pages, spending just shy of 10 minutes on my site.
I’ve tidied up my ‘world’s best logo designers?‘ article, since first publishing it, mainly because I know it attracts a fair bit of attention via Wikipedia, and I want that page to reflect well upon the rest of my content. Had I not known that this traffic was flowing to that specific page, I might well have left it alone.
The Wikipedia traffic is in contrast to visitors coming from StumbleUpon, who come and go in an average of 2 minutes 40 seconds, browsing just over two web pages in that time. Wikipedia brings people who are interested in logos, and I know a thing or two about logo design. In fact, only people arriving from Tara’s Graphic Design Blog spends longer on my site than from Wikipedia.
Why do people from Tara’s blog spend the most time on my site?
- Both blogs are related (about graphic design)
- We’ve been discussing design together for some time now
- There’s an interaction that flows between our readers
- Tara is also in the UK, which might help somewhat
- Visitors may have read a comment I left there, prompting a more curious visit
This is a short insight into just one section of Google Analytics. I’ll follow this up soon by checking other relevant statistics and showing you how they compare.
Do you use Google Analytics?
I’ve a lot to learn about using Google Analytics, and no doubt some of you have a far greater knowledge. I’m curious, do you use Google’s free stat package, and what actions do your stats prompt?
Related posts on this site
40 spot-on reader comments to “Google Analytics: referring sites”
What are your thoughts?
Simply fill in the form below. All comments are moderated so you may experience a short delay before your comment appears. Comments should be respectful of other voices in the discussion. I reserve the right to edit or delete comments at my discretion.
Hi David, many thanks for the links and of course for the traffic I receive from your site too. :)
Yes, I use analytics. But I must be honest I do try to make use of a variety of sources and methods to collect data.
I like to ensure that I don’t hinge my view of success on one view. I like a rounded look.
Using blog aggretion sites here in South Africa such as http://www.amatomu.com and http://www.afrigator.com help alot as well as analytics and other free stat tracking sites.
I find the diversity relevant and helpful.
Google has definitely outdone most competition with their display and wide variety of data avialable. It is most definitely an invaluable source of information.
I’ve just started using Google Analytics, I’ll keep what you’ve said in mind.
Why are jpegs best? I use a fair number of pngs because I use simple images that compress better as pngs.
Nic,
I agree that it’s good to use a variety. For instance, Google Analytics isn’t great for giving a very quick snapshot. This can work to my advantage, however, because in the past I think I was guilty of over-checking my stats. Google’s depth means I only look at it now and again.
I do have the SlimStat Plugin for WordPress installed, and that gives a very quick insight, right there on your WordPress dashboard.
Kristarella,
.jpg files are best for photos, due to the tonal range. .gif files are best for text, for images with large areas of flat tones, and for transparencies. I can’t comment too well on .png files, only to say that Internet Explorer 6 doesn’t work well with .png transparencies.
Perhaps you’re absolutely right to use .png’s for better compression. I’m really not sure.
I do use them, but as a newbie I’m still learning how to analyse them in order to put this data to the best use. Sometimes I find it confusing. And yep, I find a lot of great info on Dosh Dosh!
Thanks for this David excellent post, I will be taking a look at Google Analytics and see how it can help my site.
Lovin’ your work.
Thanks for the details on Google Image Search. I’ve had sites which get a fair bit of traffic from there as well, but I’ve never really looked up how it worked. :D
I use Google Analytics as well, and it’s just brilliant! Some of the stats are just nice to know (Visitor counts etc.), but there are a lot of them I use to improve my site, in particular, I like looking at bounce rates, popular content and referring sites/searches.
That’s only the basics of it though. If you can work out how to use the “goals” for a blog, then please tell, because I haven’t a clue!
I use Google Analytics - but I must admit I’m fairly new to them. Thanks for writing up the additional information regarding them in this post. Now to see if I can figure things out. Thanks.
Ok, I see what you mean for the pics. I agree jpg is best for most photos.
I think that the difference between gifs and pngs is nominal for webpages. Ten years ago browser support for pngs was patchy, but now it’s pretty good. Yes, IE6 doesn’t do transparency well and I usually give pngs a solid background for that reason, but I think it’s worth using pngs instead of gifs just for alpha transparency (during creation) and it can handle up to 32-bit colour, where as gifs only do 8-bit.
Maybe photoshop and other programs have better indexing, but I find when I make gifs in GIMP the indexed, rather than RGB colour, has a very pixelated effect, everything gets jagged. Some people make great gifs though, so I don’t know what’s wrong with mine :P
Hi David,
I also use analytics. However I don’t have the depth of info you do (much less traffic). None the less it is very interesting. It is great to see some of the strange places you get traffic from.
I also use the same tools as Nic. Amatomu.com is really great for quick and dirty stats. But I have never found the stats from Afrigator particularly useful.
In many ways Google put pay stats providers to shame. And they offer their stats free!
Nice article, just added you to my bookmark.
I’ll need to do some swotting up on the ‘goals’ for stats. Let me see what I can do though, and I’ll let you know.
Malok, Simon,
You’re both very welcome.
Kristarella,
Thanks for the extra info. You’re more clued in on the file type variations than I am, so your thoughts are appreciated.
Hi Adrian,
Absolutely - there’s no need to pay when you have Google for free! Perhaps with much larger websites, a paid service might be useful. For me, Google provide more than enough.
I use SlimStat plugin to quickly check my stats in Wordpress, and once in awhile I open my Google Analytics to get a more detailed view of my stats.
Your Stumbleupon traffic is quite impressive. How many of your pages got “stumbled” to make the total of 16,113 in a month?
Wikipedia traffic is nice too, not getting that one… at least, not yet :-)
WOW, stumbleupon.com referred so many visiters to your site! I haven’t used it until now, but after read your post, i want to try this community. I have used Google Analytics one year, i frequently cost more time on keywords which referred to my site.
SlimStat’s a nice Plugin, for sure. You know, I’m not really sure how many of my pages got ’stumbled’, but I can say that the most popular ’stumbles’ are those with imagery, or video.
Keep up the great writing and Wikipedia will be next off your checklist.
Robin,
I’m sure you’ll find StumbleUpon a great site. If only for finding new and relevant info that you wouldn’t have otherwise.
thanks, David. That’s very nice of you to say so.
I use Google Analytics regularly to analyze trends in visitors and referrals. However for real-time analysis I have switched to Clicky. Especially when you want to optimize by investigating or following individual visitor paths this package is excellent! Google Analytics is super for the broader picture and a more advanced analysis over longer terms.
That sounds really good. Have you ever tried stat counter? They have a lot of interesting stats too
I’m not familiar with Clicky, but thanks for letting me know. I’ll take a look.
Crystal,
Stat counter’s another I haven’t used. Thanks for dropping in.
Interesting that the Swiss_Mouse should stumble onto this article. He was going to write a post about his confusion between the three web stat tracking mechanisms he is using.
The Swiss_Mouse uses: Google Analytics, StatCounter and 1and1’s web statistics.
Since “going live” with the Cheddarblog, the Swiss_Mouse hasn’t been able to make heads or tails (sic) of his stats. Each of the services report completely different numbers. If you were to believe StatCounter, no one visits Cheddarblog. If you look at 1and1, it says there are plenty of unique viewers and pageviews. Google Analytics is …..in between.
The Swiss_Mouse is inclined to believe EACH of them is “partially right”, but he has yet to interpret the results entirely.
Thanks for the post.
Hi Swiss_Mouse I have the same issue, I use Google Analytics, 1and1 and MyBloglog stats and all are very different.
That’s interesting to know there’s such a difference between the packages you use. Should you take a middle ground, or is it better to believe Google over the lesser trusted offerings? It’s an interesting question.
If there’s one thing I don’t like about Google Analytics, it’s that you can’t get a quick snapshot, but of course there are other places you can, such as the packages you mention.
Tara,
I used to check MyBlogLog stats quite regularly, but since removing the widget, I no longer have access. They were very user-friendly.
Yes what’s great about the MyBlog Log stats is that it actually shows you what key words were searched for to arrive at your site which is pretty useful, can you do this with Google Analytics - I haven’t seen that feature?
Here’s a review from Pronet Advertising which you might find interesting.
Thanks I’ll take a look at that
The Swiss_Mouse wonders if Google is “smart enough” to weed out bots. That might account for some. However the Swiss_Mouse’s stats seem to be more accurate at StatCounter.com
And StatCounter does show you how people got to you via keywords, etc.
WHICH is how the Swiss_Mouse found out that he has incorrectly called Annualcreditreport.com, Annualcreditreports.com. People were finding cheddarblog via the incorrect spelling!!!!
The Swiss_Mouse will post an artile comparing his stats after the 1st of the month.
Cheddar to all!!!
Tara, Google Analytics can show you keywords - check under the menu on the left of the page, under Traffic Sources. There’s quite a bit more than meets the eye if you invesitgate that menu (shame it’s so grey and un-noticeable). ;)
Thanks Kristarella, I will take a look.
I’m a stumbleupon user and I use it quite frequently. Stumble reminds me of a remote controlling the TV. Just flipping from channel to channel. I think the fact that stumbleupon users spend an average time of 2:40 on your blog is pretty good when you think about it. Lots of its users go from site to site very quickly.
Analytics is nice for the average user, but if you’re doing any real marketing within Google then I really don’t see why anyone would use it, especially if you’re trying to rank for a term in Google. Another thing that baffles me is people using it for sites they advertise in AdWords, just doesn’t make any sense. You’re giving out info to Google, and if your site(s) are something special then believe me, Google uses that data ;)
Of course, you make a great point, that Google uses the data to rank your site accordingly. I think the point of using your Analytics is to see what traffic sources are of most value. Then you can maximise those sources with partnerships / affiliations etc. At least that’s one simple way of using the info.
Hi, David. Excellent article, thanks for the great information as usual.
A question for you….
When I read articles from your site using an RSS reader like Newsgator and never actually come to the site, does that show up somewhere in the statistics Google Analytics provides?
Depending on your answer, I may have a second question….
Thanks!
Dale
You know, I think that the Google stats record the visit from your feed reader, only to access the article. You could simply mark all your reader posts as ‘read’ and close your browser, but since your reader pulled the info off my site it counts as a hit / visit (not sure which).
If you think I can help with another question just fire away.
So there would be no referrer in this case….
Hmmm… I guess I had imagined that companies like Newsgator pulled a copy of your feed then served it to requestors … meaning, in effect, many people could read the articles off Newsgator’s single access to your site. I thought that because I sometimes see delays between the appearance of an article on a site and the available of that article via a reader. Maybe I’m over-analyzing, but I was actually trying to figure out if I’m short-changing anyone of hit counts (or of anything else) by reading their articles via Newsgator rather than going directly to their sites. Sounds like you don’t think I am….
Or is this a different subject entirely?
Thanks,
Dale
The Swiss_Mouse just posted a comparison of Page view and Unique visitors on Cheddarblog.
Interestingly, Google and Statcounter were “close” but the numbers for 1and1 Webstats were ….well…very high. Too high.
Thanks for the tip. I haven’t tried analyzing traffic by source much before. StumbleUpon is my top referrer too with 1.80 pages/visit and 1:46 average time on site. Referrals from problogger were 1:58 pages/visit and 3:41 average time on site, indicating higher quality traffic.
I’ve found some of the best referrals in terms of time spent on site are from forums, anything from 8 to 13 minutes on site.
This information will be very useful in helping to shape future posts.
Sorry it’s taken a while to respond. I’m really no expert on Newsgator. In fact I’m hardly familiar with it at all, so I can’t give the answer you’re looking for. I think that anyone you subscribe to will be very glad of your attention, but I can appreciate you wanting to give more (hits).
Swiss Mouse,
Interesting to know. Thanks for commenting.
Dean,
I agree entirely how the info is great for shaping future posts, and you also know that it’s worthwhile advertising, or contributing, in forum threads, due to the higher quality of traffic gained. Thanks for the stats.