So what seems like years ago when I was at SES San Jose one of the session I went to was on analytics (see Lisa’s coverage Issues In Analytics). Which wrinkled a few eyebrows, and got a few questions like “hey I didn’t know you were interested in analytics”, however I was reminded yesterday of why you need to have a basic understanding of how and why these things work.
I’ve been involved in analytics long enough to know it’s an inexact science and an estimate at best. I’ll show you what I mean. I recently had a story go viral and hit the hot/popular/homepage on several social sites, here are the results
Google Analytics – 8,000 visitors and 14,000 page views
Log file Analytics via clicktracks- 8.500 visitors and 20,000 page views
CrazyEgg – 7,500 visitors and 15,000 page views
Using goal tracking for feeds Google analytics shows 4 people clicked on the RSS subscribe button, however feedburner shows an increase of 57 new subscribers. I get that some people will subscribe from the RSS in the address bar but there should be a closer correlation IMHO …
Popularity: 18% [?]
Related posts:
- Google Analytics iPhone App Review I was aske
{ 13 comments }
Hey I know what you mean. I recently built an in house traffic analysis deal on one of my sites and have just been using that for now. With each pageview a new row is popped into a nice little mysql database so I know exactly how many page views there are.
Thats one of the problems with using multiple analytics solutions. An old adage applies here: A man with one watch always knows what time it is. A man with two watches is never sure.
Michael, when somebody creates a google personal homepage and types in “seo” it automatically adds your feed.
I’m guessing most of your 57 subscribers came from there.. which would track in feedburner but not in analytics.
@ryan it wasn’t on this blog it was from somewhere else. The site was brand-newsih so the likelyhood of people finding it without the social story is highly unlikely
Yeah well, tell that to my boss!
Kevin – I don’t wish to appear mean but I’d rather have the mean of many than trust in one.
@Ro: just give him the link.
I always use the icon in the address bar to subscribe to feeds.
Just a quick commentary on your stats discrepencies:
-Log file analysis (ex clicktracks) will always show higher number because (bot)”noise” is included, as well as non-image-javascript-or-’stats-block’ users.
-With crazy egg and google it is highly likely to be a question of how they define a unique visitor. (cookie, IP address, combination of factors, etc). Of course the small difference in hits is likely just server load speed. (and possibly a reporting delay.
-Of course crazy egg is really a click-heat map program, so comparing it for stats isn’t always the best idea.
-In the end you could always use GoStats as a solid authority. Note: GoStats will break down the different formats of traffic so you know what you are getting.
-email me if you’d like a further explaination.
I always only focus on my analytics through AWStats mostly. The other stats programs are too vague although I do like ClickTracks. I’ll test out Gostats immediately — I may create an article on analytics then and compare the top software suites for the best accuracy and results. Does anyone know of any big comparison sites for analytics??
>Does anyone know of any big comparison sites for analytics?
*(but new features on GoStats have been released nearly monthly and many major ones are expected soon – so I don’t blame the timeliness of tellertest)
http://tellertest.com/en
-Mind you, their GoStats review is a little outdated
Yah, comparing stats from different analytics programs is like comparing apples to oranges. My suggestion is to pick one and stick to it. I’ve found Google Analytics to be the easiest to understand – but if you’re looking at stats for a very large website, I would suggest Omniture.
PS. If anyone is interested, Search Engine Land is working on a report that will compare many of the popular analytic packages. See the interim report here
http://www.stonetemple.com/articles/analytics-report-may-2007.shtml
Comments on this entry are closed.